Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Majoring in music: how conservatory training changes the brain
(USC Thesis Other)
Majoring in music: how conservatory training changes the brain
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
Majoring in Music:
How Conservatory Training Changes the Brain
Meghen Miles Tuttle
Drs. Antonio and Hanna Damasio, Advisors
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
May 2014
© 2014
Meghen Miles Tuttle
All rights reserved.
i
To my parents,
Who taught me to reach for the stars,
And to Jay,
Who helped get me there.
ii
ACKNOWLEDGEMENTS
I would like to express my deepest gratitude to my mentor and committee co-
chair, Dr. Hanna Damasio. You took me under your wing, helping me to discover my
love for neuroanatomical research, and the methodology with which I would approach
questions of music and the brain. Without your guidance, encouragement, and friendship,
I would not have finished this dissertation.
To my advisor and committee chair, Dr. Antonio Damasio, I offer my heartfelt
thanks for your frank advice, for allowing me to use the Brain and Creativity Institute’s
beautiful Cammilleri Hall for my dissertation defense, and for all of the wonderful
opportunities afforded me over the years I spent as a graduate student in the BCI, both
scientific and artistic. It is my hope that you and Hanna know how deeply I have
appreciated this time, and the chance you took on me.
To my committee member, Dr. Laura Baker, I thank you for your years of support
and guidance. I have not forgotten the encouragement towards research you gave me
while I was still an undergraduate student, taking your class. Thank you for giving me a
place in your lab, and for being both my friend and advocate.
To my outside committee member, Dr. Robert Cutietta, I thank you for all of your
thoughtful commentary and support. I have truly enjoyed having an outside member who,
as both musician and skilled researcher, took the time to be so involved in my process.
To my labmates and other colleagues, both at the BCI and within the
Neuroscience Graduate Program, I will be forever grateful for your collaboration,
support, idle chit chat, and inspiring suggestions. In particular, for your help with
iii
preparations for my oral qualifying exam and dissertation defense, I thank Helder Filipe,
Glenn Fox, Kingson Man, Dr. Xiaofei Yang, Dr. Assal Habibi, and Andrea McColl. Your
constructive criticism turned what could have been truly mediocre talks into shining
moments for me.
To everyone who helped me with data collection and analyses, I am in your debt.
I would like to thank Dr. Jiancheng Zhuang for help with choosing scanning protocols
and assistance with MRI data collection at the Dornsife Neuroimaging Center. I would
like to thank Dr. Jonas Kaplan for help with everything FSL, needed in particular for my
VBM analyses. Finally, I would like to thank So-Young Choi, Dr. Jessica Wisnowski,
Dr. Dimitrios Pantazis, Dr. Anand Joshi, Dr. Justin Haldar, Dr. Chitresh Bhusan, and Dr.
David Shattuck for assistance with my BrainSuite analyses.
I would like to acknowledge the deep support given to me by my friends and
family, most especially my parents, The Very Rev. Dr. Sir Richard and D’aun Miles. You
have always encouraged me to chase my dreams, and taught me that with hard work,
anything is possible. Your unconditional love has been my buoy.
Finally, I offer my very special thanks to my husband, Jay Tuttle. Your
unconditional love, encouragement, and vigilant nudging saw me through to the finish
line.
This work is supported by the Brain and Creativity Institute, the USC
Neuroscience Graduate Program, and the USC Graduate School.
iv
LIST of FIGURES
CHAPTER 3
Figure 1: Architecture freshmen as compared to architecture seniors in
precentral gyrus (p=0.428*). 78
Figure 2a: Architecture Seniors > Voice Majors (first run), p=0.0582 98
Figure 2b: Architecture Seniors > Voice Majors (second run), p=0.0976 98
CHAPTER 4
Figure 1: The three surfaces generated by BrainSuite, marked sulci, and the final output,
a labeled pial surface 107
CHAPTER 5
Figure 1: VBM results in music seniors when compared to music freshmen
(p=0.014*) 171
Figure 2: BrainSuite results in music seniors when compared to music freshmen in left
postcentral gyrus 175
Figure 3: VBM results in music majors when compared to architecture majors
(p=0.0362*) 176
Figure 4: BrainSuite results in music majors when compared to architecture majors in
left Broca’s area 176
Figure 5: BrainSuite results in piano majors when compared to strings majors in left
precentral gyrus 177
Figure 6: BrainSuite results in piano majors when compared to voice majors in right
precentral gyrus and right postcentral gyrus 177
Figure 7: BrainSuite correlation results in instrumentalists in left precentral gyrus 178
Figure 8: BrainSuite correlation results in vocalists in left pars triangularis, left Heschl’s
gyrus, and bilateral cingulate gyrus 179
Figure 9: The two most immediately relevant implications of musical experience-
dependent neuroplasticity research 182
v
LIST of TABLES
CHAPTER 2
Table 1: Population Characteristics, by Major and Year 56
Table 2: Population Characteristics, Architecture Students with Significant Musical
Training (AWM) 57
Table 3: Population Differences, by Major 58
Table 4: Population Differences, by Major and Year 59
Table 5: Population Characteristics, Instrumentation 62
Table 6: Population Characteristics, by Instrument Group (Music Majors) 63
Table 7: Population Differences, by Instrument Group (Music Majors) 64
CHAPTER 3
Table 1: Areas of increased gray matter density found via voxel-based morphometry
analyses of musicians as compared to nonmusicians. 69
Table 2: Voxel-Based Morphometry Results, Planned Analysis (Major and Year) 73
Table 3: Voxel-Based Morphometry Results, Exploratory Analysis 1
(Instrumentation) 81
Table 4: Voxel-Based Morphometry Results, Exploratory Analysis 2 (Gender) 86
Table 5: Voxel-Based Morphometry Results, Exploratory Analysis 3 (Architects with
Significant Musical Experience (AWM)) 94
CHAPTER 4
Table 1: BrainSuite Results, Planned Analysis (Major and Year) 114
Table 2: BrainSuite Results, Exploratory Analysis 1 (Instrumentation) 119
Table 3: BrainSuite Results, Exploratory Analysis 2 (Architects with Significant Musical
Experience (AWM)) 123
Table 4: Comparisons Driven by Outliers, Exploratory Analysis 3 124
Table 5: Correlational Analyses, by Measurement 129
Table 6: Planned Analysis, by Major and Year 136
Table 7: Planned Analysis, by Year within Major 143
Table 8: Exploratory Analysis 1 (Instrumentation), by Three Groups (Instrumentalists,
Vocalists, Architecture Majors) 150
Table 9: Exploratory Analysis 1 (Instrumentation), by Instrument (Piano Majors, Strings
Majors, Voice Majors) 155
Table 10: Exploratory Analysis 2 (Architecture Majors with Significant Musical
Training (AWM)) by Three Groups (Architecture Majors, AWM Subjects,
Music Majors) 160
vi
ABSTRACT
Studies investigating musical experience-dependent neuroplasticity clearly show the
presence of neuroanatomical differences between musicians and non-musicians, in a
variety of regions having functionally to do with music perception and performance,
including primary motor cortex (precentral gyrus), somatosensory cortex (postcentral
gyrus), primary auditory cortex (Heschl’s gyrus), inferior frontal gyrus, cerebellum, and
corpus callosum. The current study sought to answer the question of whether or not four
years spent in conservatory-style training would be long enough to elicit noticeable
changes in these regions of interest (ROIs). A comparison group of architecture majors
was chosen for the educational similarities between architecture and music performance
degrees; previously published studies have largely compared musicians to non-musicians
of many disciplines. Two main methods were used to analyze the dataset collected:
voxel-based morphometry (VBM), and a semi-automated anatomical approach using the
software BrainSuite. The two methods differ in terms of time required for analysis, the
types of comparisons that can be made, and what they measure. Regarding the latter,
VBM measures groupwise differences in grey matter density, and BrainSuite analysis
yields volumetric and thickness measurements in individual subjects. The results were as
follows: 1. Music seniors exhibited increased grey matter density in right cerebellum, and
total volume, grey matter volume, and white matter volume increases in left postcentral
gyrus, when compared to music freshmen. 2. Music majors as a whole exhibited
increased grey matter density in left anterior insula, and total volume and grey matter
volume increases, as well as a statistical trend toward increased white matter volume, in
left Broca’s area, when compared to architecture majors as a whole. 3. An effect of
vii
instrumentation (piano, strings, or voice) could be seen in bilateral precentral gyrus and
right postcentral gyrus. 4. Finally, the age of onset of training on the primary instrument
was found to negatively correlate with measurements in left precentral gyrus, left pars
triangularis, left Heschl’s gyrus, and bilateral cingulate gyrus, regions found in both the
extant literature and my hypotheses for this study. These regional correlations differed
between instrumentalists and vocalists. I conclude that studies undertaken in the field of
musical experience-dependent neuroplasticity have possible implications in the fields of
education and music therapy.
viii
TABLE of CONTENTS
DEDICATION i
ACKNOWLEDGEMENTS ii
LIST of FIGURES iv
LIST of TABLES v
ABSTRACT vi
CHAPTER 1 1
A. INTRODUCTION
B. BACKGROUND 2
i. Relevant Neuroanatomical Studies on Musicians 2
1. Manual Volumetric Analyses 3
Corpus Callosum 3
Motor Cortex 5
Cerebellum 8
Planum Temporale 8
Inferior Frontal Gyrus 9
2. Grey-Matter Density (Voxel-Based Morphometry (VBM)) 10
Adults 10
Children 18
Keyboardists 20
Mixed Instrumentation 20
3. White Matter Volume (Diffusion Tensor Imaging (DTI)) 21
4. Summary 26
Corpus Callosum 27
Motor Cortex 27
Somatosensory Cortex 28
Cerebellum 28
Heschl’s Gyrus 28
Inferior Frontal gyrus 29
ii. Influence of Genetics 30
iii. Non-Musical Experience-Dependent Structural Plasticity 34
1. Learning and Memory 34
2. Motor Skills 37
3. Repetitive Magnetic Stimulation (rTMS) 39
4. Summary 39
iv. Possible Microscopic Mechanisms 40
1. Histology 40
2. Neurogenesis 41
3. Synaptogenesis 42
4. Axonal Sprouting 43
5. Other 43
v. Summary 44
ix
TABLE of CONTENTS
CHAPTER 2 46
A. INTRODUCTION 46
B. SPECIFIC HYPOTHESES 50
C. PREDICTIONS 50
D. SUBJECTS 51
i. Specific Populations 51
ii. Population Characteristics 53
1. Population Characteristics by Major and Year 54
2. Population Characteristics by Instrument Group 60
CHAPTER 3 68
A. BACKGROUND 68
B. METHODS 70
i. Subjects 70
ii. Image Acquisition 70
iii. Analysis 71
C. PROPOSED/ORIGINAL ANALYSES 72
i. Results 72
a. Stage 1 72
b. Stage 2 73
ii. Discussion 74
D. EXPLORATORY ANALYSES 80
1. Instrumentation 80
i. Results 80
a. Stage 1 81
b. Stage 2 82
ii. Discussion 83
2. Gender 86
i. Results 86
a. Stage 1 87
b. Stage 2 87
ii. Discussion 88
3. Architects with Significant Musical Training 93
i. Results 93
a. Stage 1 93
b. Stage 2 93
ii. Discussion 95
E. GENERAL DISCUSSION AND CONCLUSION 95
i. The Study-Specific Template 97
x
TABLE of CONTENTS
CHAPTER 4 101
A. BACKGROUND 101
B. METHODS 104
i. Subjects 104
ii. Image Acquisition 104
iii. Analysis 104
1. Structural MRI Data 104
2. Diffusion Data 108
3. Statistical Analysis 109
C. PROPOSED/ORIGINAL ANALYSES 111
i. Results 111
ii. Discussion 113
D. EXPLORATORY ANALYSES 117
1. Instrumentation 117
i. Results 117
ii. Discussion 119
2. Architects with Significant Musical Training 121
i. Results 121
ii. Discussion 125
E. CORRELATIONAL ANALYSES 126
i. Age of onset of use of the first instrument 127
ii. Age of onset of use of the primary instrument 128
iii. Intensity of practice on the primary instrument 128
iv. Discussion 128
F. GENERAL DISCUSSION AND CONCLUSION 132
CHAPTER 5 168
A. DISCUSSION: METHODOLOGY AND SAMPLE 168
B. SUMMARY 174
C. FUTURE DIRECTIONS 180
D. IMPLICATIONS 181
REFERENCES 184
1
CHAPTER 1
A. INTRODUCTION
There is little doubt that musical training changes the way one thinks about music. There
is also little doubt that musical training alters both the structure and function of the brain
via mechanisms underlying neuroplasticity, at least in the developing brain
(Schellenberg, 2003, Schlaug, 2006). Neuroanatomical studies of musicians have been
undertaken on both adults and children. These studies have produced inconsistent results,
likely due to differing subject populations (gender, age range, and instruments played)
and/or limitations of the analysis techniques used. Reliable differences in professional
musicians at the adult level, some of which can be corroborated by results in children
taking music lessons, have been found, not surprisingly, in motor and auditory related
areas as well as the inferior frontal gyrus. But the gap between childhood music lessons
and professional musicianship has not been addressed. What remains to be established is
(a) if intensive conservatory level training can induce further structural changes in the
musician’s brain; and (b) if these changes are distinguishable from those formed in other
university students undergoing similarly rigorous but nonmusical artistic degree
programs. Herein detailed is a study conducted to address these questions.
Entering freshmen music majors (string players, keyboard players, and vocalists)
were compared in terms of volumetric differences in grey matter and white matter, with
senior music majors (same instrument groups). Both cohorts of conservatory musicians
were compared with age, gender, and university level matched non-musicians pursuing
architecture majors. Volumetric brain data were collected in a 3 Tesla Siemens MRI
scanner. Both T1 and diffusion scans were obtained. Several techniques were used to
2
analyze the data, including one automated method, voxel-based morphometry (VBM),
performed in FSL. Statistical analyses were performed in SPSS on all data collected
including a comprehensive performance background check on each subject. The goal of
this study was to help determine the potential role of intense conservatory training - with
its historical, aural, and theoretical analysis demands - in shaping the structure of
musicians’ brains.
B. BACKGROUND
i. Relevant Neuroanatomical Studies on Musicians
Results from previous structural comparisons between musicians and non-musicians have
sometimes been contradictory. These studies have incorporated a variety of imaging and
analysis techniques including volumetric analyses performed using manual tracing
techniques or automated procedures such as voxel-based morphometry (VBM), and
diffusion tensor imaging (DTI) analyses. Each technique brings its own set of limitations,
likely contributing to the mixed nature of the results. There has been no study yet to date
to incorporate all three of the above techniques, although a few studies have utilized a
combination of two of the three techniques, to their advantage (Schneider et al, 2005a,b,
Han et al, 2009, Scholz et al, 2009). A review of the studies falling into each category
follows.
3
1. Manual Volumetric Analyses
Volumetric analyses using manual tracing tend to be time-intensive due to the typical
employment of hand-tracing methods. Thus, subject populations and regions of interest
tend to be relatively small and focused, compared to studies using automated methods.
For the purposes of this review, studies will be addressed in relation to regions of interest.
Corpus Callosum
Two studies have investigated the size of the corpus callosum in musicians
compared to non-musicians. The first, published by Gottfried Schlaug and colleagues in
1995, found that musicians have a significantly larger anterior corpus callosum when
compared to age, sex, and handedness-matched controls. Schlaug’s group used manual
tracing on 60 subjects (30 musicians and 30 non-musicians), all of whom were university
students or recently graduated. The non-musicians were all medical students or young
faculty. The majority of subjects were men, although both sexes were included. Anterior
corpus callosum size was significantly larger in musicians (age of onset of musical
training was earlier than 7 years of age). Schlaug’s group interpreted their results in light
of animal study results indicating a link between larger corpus callosum size and both
increased interhemispheric connectivity and a larger degree of hemispheric symmetry
(reviewed in Schlaug 1995). This study was performed on keyboard and strings players;
it remains to be seen if the findings extend to musicians who specialize in other
instruments.
4
The second study on corpus callosum size in musicians was also conducted by
Schlaug’s group (Lee et al, 2003). This study again found increased anterior corpus
callosum size in musicians (n=56, keyboard and/or strings players) compared to non-
musicians (n=56). The groups were evenly split by gender, and again, all were currently
or recently graduated university students. Manual tracing was employed on the corpus
callosum. Interestingly, in this study a gender difference was found; male musicians
showed the expected increased anterior corpus callosum size compared to non-musicians,
while females showed no statistically significant difference based on musicianship. In
fact, the female musicians in this study actually showed a trend toward a decrease in
corpus callosum size (both anterior and posterior) when compared to female non-
musicians. Schlaug’s group interpreted these data in light of previous human studies
suggesting a higher degree of hemispheric symmetry in right-handed females when
compared to handedness-matched males. The authors referenced other studies suggesting
that hemispheric asymmetry is inversely related to corpus callosum size; thus, one might
expect larger callosal size in more symmetric brains, or smaller callosal size in less
symmetric brains. Entrainment of the nondominant hand could alter hemispheric
symmetry; in females, who typically have a higher degree of symmetry, the effect of this
entrainment could be a decrease in symmetry. The authors suggested that the trend
toward a decrease in anterior callosal size in female musicians could be explained by the
bimanual nature of keyboard and/or strings playing, which entrains the nondominant
hand, thereby decreasing hemispheric symmetry in female brains (Amunts et al, 2000,
Lee et al, 2003). Additionally, Schlaug’s group did not account for absolute pitch (AP),
which was present in their female musician population to a significantly higher degree
5
than in their male musician population. AP has been linked to asymmetry in musicians,
particularly in the planum temporale (Keenan, 2001). The high incidence of AP
combined with the bimanual nature (most importantly, the use of the nondominant left
hand) of keyboard and strings playing, therefore, could have led to a lesser degree of
hemispheric symmetry in the brains of their female musician population (who, according
to their theory and the evidence presented in Amunts, 2000, would have begun with more
symmetric brains than the males (despite right hand dominance); thus, training of the
nondominant hand would increase asymmetry rather than the symmetry represented in
the male musicians) when compared to the female non-musicians, yielding smaller
corpus callosum size, due to the inverse relationship between hemispheric asymmetry and
corpus callosum size (Lee, 2003). The chief problem with this argument lies in the fact
that the hand area of the corticospinal tract does not cross in the corpus callosum; rather,
it passes through the posterior limb of the internal capsule and should not have an effect
on corpus callosum size. Their argument only works for general hemispheric asymmetry.
Motor Cortex
Another two studies have investigated aspects of motor cortex size and symmetry in
musicians compared to non-musicians. The first, published by Amunts and colleagues
(including Schlaug) in 1997, looked at the length of the posterior wall of the precentral
gyrus, as it borders the central sulcus; they called this the intrasulcal length of the
precentral gyrus (ILPG). Length of the ILPG was calculated in the following manner:
“The ILPG was measured in a dorso-ventral sequence of horizontal sections as the length
of the contour line of the posterior bank of the precentral gyrus from the deepest point of
6
the sulcus to a lateral surface tangent which connected the crests of the pre- and
postcentral gyrus” (Amunts et al, 1997). The ILPG was cited as an index of motor cortex
size. Manual tracing was employed. Subjects were male keyboardists (n=21), and male
non-musicians (n=30), and all were current university students or within five years of
graduation. Non-musicians showed an expected pronounced leftward asymmetry of the
ILPG, while musicians showed more hemispheric symmetry in the ILPG. The most
pronounced between-groups differences were in the most dorsal part of the hand knob
region of the ILPG, particularly in the right hemisphere. Amunts’ group also found that
ILPG size correlated negatively with age of onset of musical training, meaning that the
younger a subject began lessons, the larger their adult ILPG, particularly in the right
hemisphere (Amunts, 1997). Although the results of this study are interesting, the use of
the ILPG rather than motor cortex size is questionable. Further studies need to be
performed on the motor cortices of musicians of both sexes, and both those with highly
bimanual instruments, as well as those playing instruments requiring more usage of the
right or left hands.
The second study, published by Bangert and Schlaug in 2006, looked at hand
knob size and hemispheric symmetry in keyboard players (n=16), strings players (n=16),
and non-musicians (n=32). Subjects were current university students or recent graduates.
Bangert and Schlaug carefully excluded potential subjects who played both keyboards
and strings, a feat which would be difficult at most conservatories. Hand knobs were
manually traced by five observers. Results yielded a leftward asymmetry of the hand
knob in keyboardists, a rightward asymmetry of the hand knob in strings players, and a
relative symmetry of the hand knob in non-musicians, despite the fact that all subjects
7
were right-handed. Bangert and Schlaug’s results suggest that the manual requirements of
different instruments yield differential brain adaptation (Bangert & Schlaug, 2006). The
hand knob symmetry of the nonmusician group leads one to wonder if these non-
musicians were highly skilled at some other bimanual task, such as typing or video game
play. The leftward asymmetry in keyboardists can be explained by the fact that right-
handed keyboardists, besides using their right hands heavily outside of keyboard playing,
often have more complicated note patterns to play with the right hand. Rightward
asymmetry in strings players can be explained by the fact that strings players (particularly
violin and viola players) will most often use their right hands only to hold and move a
bow, while their left fingers move deftly across the strings. The nonmusician results from
this study fit with the results of a recent investigation into handedness and hand knob
symmetry/asymmetry (Allen, 2007). J. Allen reveals that, despite some reports in the
literature of leftward asymmetry in primary motor cortex (Amunts, 1996), the general
right-handed population exhibits either symmetry or a slight rightward asymmetry of the
hand knob area of primary motor cortex (Allen, 2007). Although one might posit that
right-handed non-musicians should show a leftward asymmetry of the hand knob (due to
increased usage of the dominant right hand over the left hand), the above nonmusician
results, in combination with the results of Allen’s review, suggest the possibility that non-
musicians are skilled at some other type of bimanual task that may have trained their non-
dominant left hands, yielding symmetry in their hand knobs. Replication of this study
with subject background information obtained for other types of bimanual tasks would
help clarify the issue. The addition of an instrument group requiring a greater right vs.
left hand dexterity (our addition of brass players), should also yield a clean comparison
8
for the keyboard and strings groups, with, presumably, a pronounced leftward asymmetry
of the hand knob area.
Cerebellum
Hutchinson and colleagues (including Schlaug) published a study in 2003 on
cerebellar volume in professional keyboard players (n=60; 30 male, 30 female) compared
to non-musicians (n=60). Cerebellums were manually segmented and traced. Results
showed that male musicians had significantly greater absolute and relative cerebellar (not
total brain) volume compared to non-musicians. This greater volume positively correlated
with lifelong intensity of instrumental practice. As with the aforementioned anterior
corpus callosum study (Lee, 2003), female musicians showed no absolute or relative
cerebellar volume differences when compared to non-musicians. Hutchinson’s group
discussed the lack of a female cerebellar size difference in light of the fact that women in
general tend to have larger relative cerebellar volumes than men (Hutchinson, 2003).
Both female cohorts showed relative cerebellar volumes of roughly the same size
(although very slightly larger) as the male musician group. The gender difference, as well
as the correlation with lifelong degree of intensity in practice (for men), is intriguing, and
would benefit from replication.
Planum Temporale
Keenan and colleagues (including Schlaug) published a study in 2001 on the
planum temporale (PT) and its anatomical relationship to absolute pitch in musicians. As
they discussed in the report, AP likely results from a combination of genetic influence
9
and early exposure to musical training. Subjects for the study were AP professional
musicians (n=27), non-AP professional musicians (n=22), and non-AP non-musicians
(n=27). Both sexes were represented roughly evenly in all three populations; slightly
more females than males were studied in the AP musician and non-AP non-musicians
groups, and slightly more males than females were studied in the non-AP musician
group. Musical training had started before age 7 in both musician populations. The
planum temporale was manually identified and traced. Results revealed a leftward
asymmetry and a significantly smaller rightward absolute size of the planum temporale in
AP musicians compared to both control groups. Keenan’s group, like in most published
studies, used only keyboard and strings players. They found no effect for gender. This
well performed study would also benefit from replication. Its specificity to the AP
population makes it relevant to only a subset of musicians, although the AP/PT
asymmetry has to be taken into consideration when assembling subject groups for
neuroanatomical studies, particularly when hemispheric symmetry is to be considered.
Inferior Frontal Gyrus
In 2011, Sluming’s group published a follow-up to their 2002 study on male
orchestral musicians (reviewed in the next section) as compared to male non-musicians
(Abdul-Kareem et al, 2011). This time, the authors manually segmented the inferior
frontal gyrus into the pars opercularis (POP) and the pars triangularis (PTR). Results
supported their 2003 findings of increased grey matter in the left inferior frontal gyrus,
specifically in the left POP. No significant grey matter differences were found between
the groups in the left PTR. The use of manual techniques on individual brains can, as in
10
this study, help bolster and clarify results found using more automated groupwise
methods, such as voxel-based morphometry (VBM).
2. Grey Matter Density (Voxel-Based Morphometry (VBM))
The literature focused on specific grey matter differences between musicians and non-
musicians has produced widely disparate results, with the exception of just a few brain
regions. Whether this is due to differences in subject populations, differences in software
used for automated methods (all but one voxel-based morphometry (VBM) methods, but
some use different software versions of SPM, FSL, etc.) or a result of disparate
thresholds for statistical significance, is unclear. The widely varying regions of grey
matter differences mentioned could also be in part due to differences in nomenclature
used by separate authors to describe the same and/or overlapping brain regions.
Adults
The majority of grey matter density studies on musicians have been performed on
adult musicians, typically male keyboard and/or strings players. Gaser & Schlaug’s 2003
study yielded the highest number of brain regions with grey matter differences between
musicians and non-musicians. The authors analyzed brain images from a total of 80
subjects: 20 professional musicians, 20 amateur musicians, and 40 matched non-
musicians. All subjects were male, right-handed, keyboardists, and 18-40 years of age.
Data were processed in SPM and analyzed via fully automated VBM methods. Results
showed a positive correlation between grey matter volume and musicianship
11
(professional or amateur musician, or nonmusician) bilaterally in the following regions:
primary motor cortex, somatosensory cortex, premotor areas (no details are given),
anterior superior parietal lobe, and inferior temporal gyrus. Positive correlations with
musicianship were also found in the left hemisphere in: cerebellum, Heschl’s gyrus, and
the inferior frontal gyrus. Gaser and Schlaug found no correlation between musicianship
and white matter volume. The most interesting result in this study is the correlation of
grey matter volume increases in multiple brain regions in accordance with musicianship.
Professional musicians showed the highest amount of grey matter in all regions. Amateur
musicians showed an intermediate amount of grey matter in these areas, and non-
musicians showed the lowest grey matter volume in the brain areas mentioned above. All
the professional musicians involved in the study practiced keyboard for at least an hour
daily, and all had received formal training (conservatory training is inferred from the
description in the methods section, but not stated outright). The amateur musician group
practiced on average just over an hour less per day than the professionals, and had
received some form of formal training on a keyboard instrument (likely private lessons).
Those subjects in the nonmusician group had never played a musical instrument. The
authors interpret their widespread results by applying them as a whole to the genetic
predisposition (innate structural differences between musicians and nonmusician brains)
versus experience-induced plasticity debate; the strong positive correlation between
musicianship (professional or amateur) and volume of grey matter supports the latter
argument, although it does not rule out the former (musicians could be born with
structural differences that then become enhanced with experience). The authors
additionally discuss their lack of white matter findings as a result of the technique
12
employed; VBM is less sensitive to white matter differences than other methods (e.g.
DTI) (Gaser & Schlaug, 2003). The widespread results from this study are certainly
intriguing, but they have not been fully replicated in subsequent studies on musicians
compared to non-musicians. This study needs to be replicated in a female population of
musicians (both professional and amateur) and non-musicians in order to draw a more
complete picture of brain differences between musicians and non-musicians, as well as
conservatory trained versus amateur musicians.
Sluming and colleagues published a VBM study in 2002 that specifically focused
on grey matter density in Broca’s area of the left inferior frontal gyrus. This study
compared 26 right-handed male orchestral musicians to 26 matched controls (non-
musicians). The results revealed greater grey matter density and volume in the left
inferior frontal gyrus (IFG) in musicians than in non-musicians. Furthermore, bilateral
age-related decreases, both in dorsolateral prefrontal cortex volume and in overall
hemispheric volume, were detected in non-musicians but not in the musician group.
These findings were interpreted as suggesting greater development and/or greater
retention of grey matter by musicians over time in areas subjected to continued frequent
use. As many bilateral brain structures have been implicated as being volumetrically
greater in musicians as compared to non-musicians (reviewed here), all of these
structures (not just the left IFG focused on in this paper) could contribute to greater
bilateral hemispheric volume in older musicians as compared to older non-musicians. The
left IFG has been implicated as important to a variety of musical activities common to
orchestral musicians, including but not limited to attentive score reading, sight reading,
and the processing of musical syntax (Parsons, 2001, Sergent et al, 1992, Maess et al,
13
2001, reviewed in Sluming et al, 2002). Sluming’s group also performed
neuropsychological tests on all subjects directed at accessing general cognitive ability
and spatial ability. Although no statistically significant differences were found between
musicians and non-musicians for general cognitive ability, musicians exhibited
significantly higher spatial ability than non-musicians, perhaps related to the continued
use of spatial sound localization in an orchestral setting. Although the wide range of
instruments played by the subjects in this study suggests that these findings can be
applied to most musicians, it must, as in the Schlaug paper above, be noted that the group
was composed solely of male musicians; female musicians need to be examined before
broad conclusions can be drawn about the effects of years of training, practice, and
performance on grey matter volume in musicians.
An incredibly large study on Heschl’s gyrus (HG) in musicians and non-
musicians was undertaken by Schneider and colleagues and published in 2005 (Schneider
et al, 2005a and 2005b). The study compared HG in professional musicians (n=334),
amateur musicians (n=75), and non-musicians (n=54). Results showed a correlation
between absolute size of HG (and grey matter volume of HG) and musical ability, as
assessed by individual histories of musical training and Gordon’s Advanced Measure of
Music Audiation (a test that infers musical aptitude from auditory imagery ability). The
study also examined pitch perception preferences (fundamental vs. spectral pitch
listeners) in musicians specializing in all types of instruments, including non-
instrumentalists by trade (vocalists and conductors). There was an asymmetry in lateral
HG correlated with pitch perception preference; fundamental pitch listeners displayed
larger left lateral HG, while spectral pitch listeners display larger right lateral HG. The
14
authors were able to match listener types (fundamental or spectral pitch) with their
instruments, yielding an orchestral map of pitch perception preference (Schneider et al,
2005a, 2005b). Whether these pitch preferences (and related HG volumes) are pre-
existing and lead a musician to choose a particular instrument, or whether that instrument
and its practice yields pitch preference and HG volume changes, is unclear, and is not
really discussed. It seems odd that the only major difference between musicians and non-
musicians would lie in HG, or in auditory cortex in general for that matter. It seems more
plausible that HG was what the authors wished to focus on for this particular report, as
might have been the case for left IFG in the previously discussed Sluming paper. This
study by Schneider’s group is a good example of why diversity in instrumentation in the
population of musicians is important. It is difficult to draw conclusions about the entire
population of musicians from studies based solely on keyboard and strings players.
Bermudez and Zatorre studied, in 2005, grey matter differences between
musicians (n=43) and non-musicians (n=51). The subject selection is not fully detailed.
Results indicated larger grey matter volume in musicians than in non-musicians in the
right planum polare and planum temporale. Although the musician group included 22
musicians with absolute pitch, these subjects did not account for the difference in planum
temporale grey matter volume found in musicians; all musicians had larger grey matter
volume in planum temporale. The authors do not comment on any volumetric
asymmetries in planum temporale in the absolute pitch musicians, as would be expected
in view of Keenan’s study (2001), discussed in volumetric analyses, above. Although the
results yielded by the VBM analyses are small in terms of regions of the brain, it bears
consideration that Bermudez and Zatorre do not make their subject selection process very
15
clear. The musician group is said to have had “10 years or more of musical experience.”
This is not necessarily indicative of professional musician status, and nowhere in the
report is musical expertise discussed; these musicians could in fact simply be students
who have played instruments as an extracurricular activity since middle school. The
authors also do not mention the ages of their subjects, leaving the possibility that there
could be age-related differences, either due to development or decline, in their subjects,
accounting for the minimal differences found in musicians versus non-musicians. Both
males and females were included in this study, which could mean that the results may
have been affected by gender differences.
A 2009 study on adult musicians (keyboardists) and non-musicians conducted by
Han, combined structural MRI and DTI, looking at both grey matter and white matter in
both male and female keyboardists (n=19) and non-musicians (n=21). The musician
group, when compared to non-musicians, showed higher grey matter volume in the left
sensorimotor cortex and right cerebellum, and lower grey matter volume in the right
orbital frontal gyrus and left anterior cingulate cortex. DTI analyses showed higher
fractional anisotropy (FA; FA is interpreted as proxy to white matter volume) in the right
posterior limb of the internal capsule (PLIC), the midbrain area of the brain stem, and the
left inferior frontal gyrus. No areas of lower FA were found in musicians compared to
non-musicians. What is clear from these results is that the chief differences seen in
musicians are in areas subserving motor function and/or representation of sensory
information related to those motor functions. The authors suggest that their FA findings
in left IFG might support Sluming’s grey matter findings in the same region, which they
failed to replicate. One might recall that Sluming’s study (discussed earlier) included
16
subjects with a very wide age and instrument range; although the average age for Han’s
subjects was 22, the range was from 19-28. Also, Sluming’s subjects were male
orchestral players, while Han’s subjects were both male and female pianists, for which no
separate male vs. female analysis was reported. A downfall of Han’s study is that not all
details about the pianists’ musical background were reported. In addition to not recording
performance background information, the authors also did not record whether or not the
pianists had absolute pitch. All subjects were Chinese, which makes it more likely that a
higher percentage of the subjects may indeed have had absolute pitch, given that absolute
pitch correlates positively with a tonal language as a first language (Deutsch, 2006).
Regarding the differences in instrumentation, perhaps there is something to the
performance practice of pianists, with fewer collaborations, and typically fewer people
involved when collaborations occur (as compared to the communal performance practice
of orchestral musicians) that may fuel the white matter changes in left IFG. One would
need a detailed performance history (solo versus collaborative performances and their
regularity, as well as practice and rehearsal schedules) to justify claims about such a
finding. Alternatively, perhaps the methods used are to blame for the grey matter/white
matter disparity between these two studies; Sluming’s study had more subjects in each
group, possibly yielding a more statistically significant grey matter result in the left IFG.
Han’s group employed DTI as well as VBM, increasing the likelihood of finding reliable
white matter results. The combination of structural MRI and DTI in Han’s study yielded
more interesting results than each of the techniques alone. It can be improved upon by
replication in a more mixed group of musicians (different instrumentation), adding
gender-based comparisons and detailed individual musical histories to the final analysis.
17
Groussard used VBM to investigate hippocampal differences between musicians
(n=20) and non-musicians (n=20) in 2010. Musicians of mixed instrumentation were
recruited, and both groups (M and NM) were split evenly by gender. Region of interest
(ROI) analyses were performed within the context of VBM on areas showing a functional
difference during a musical familiarity task between musicians (mixed instrumentation)
and non-musicians. Of those ROIs (bilateral hippocampus, calcarine gyrus, orbitomedial
frontal gyrus, cingulate cortex, and superior temporal gyrus (including HG)), differences
in grey matter density were revealed in the left anterior hippocampus, a region suggested
to be correlated with verbal memory recall (Groussard et al, 2010).
A 2013 study by James et al used VBM to compare the brains of 20 professional
pianists, 20 amateur pianists, and 19 non-musicians. Both men and women were included
in the sample. The authors found increases in grey matter density in right fusiform gyrus,
right medial orbitofrontal gyrus, left inferior frontal gyrus, left intraparietal sulcus, left
Heschl’s gyrus, and bilateral posterior Crus II (cerebellum). All increases positively
correlated with expertise. Findings in left inferior frontal gyrus, left Heschl’s gyrus, and
posterior cerebellum were not unexpected, given the extant literature. The authors discuss
increases in right fusiform gyrus in light of the region’s known role in visual pattern
recognition, needed for complex musical score reading. Following the model put forth by
Janata, James suggests that the medial orbitofrontal gyrus is important for determining –
and following – tonality in music. Findings in the left intraparietal sulcus were suggested
to be related to visual-motor coordination needed to play piano and musical score
reading. Additionally, the authors report grey matter decreases correlated with expertise
18
in bilateral periolandic and striatal areas, which they explain in terms of motor skill
automation in experts (James et al, 2013).
Children
Although studies on adults offer a host of structural clues as to what makes a
musician’s brain different from a nonmusician’s brain, longitudinal studies, particularly
in children, remain the best way to elucidate the possible cause of these differences. In
other words, is it music making as an act and practice that affects neuroplasticity, or is it
an initial structural difference that causes a child to seek out and maintain the musical
process? There is really only one relevant grey matter study on children on which to
report. Schlaug and colleagues are currently conducting a longitudinal study that seeks to
shed light on the issue of nature vs. nurture through the random assignment of children to
music lessons. The ongoing study (since 2003) began by assigning 50 children (5-7 years
of age) to music lessons (two-thirds are studying piano, while one-third is studying string
instruments; this is a fairly typical distribution for young children taking music lessons).
In addition, 25 children, matched to the instrumentalists in age, socioeconomic status, and
verbal IQ, are being followed as controls. Preliminary results show an overall trend
toward increasing grey matter in the whole brain, directly correlated with the years spent
training on a musical instrument (Schlaug et al, 2005, Schlaug, 2006, Hyde et al, 2009).
Schlaug’s group is investigating these children using both functional and structural MRI.
His fMRI studies have revealed functional processing differences in response to music in
the musician group compared to the nonmusician group. These functional differences
began appearing in both hemispheres, largely in temporal auditory association areas as
19
well as the temporal-parietal junction, as fast as 12 months after the onset of training;
structural changes, as indicated by the increasing trend in grey matter over training time,
are slowly following suit. The control group has not shown any significant functional or
structural changes. A comparison of these 5-7 year old subjects to a group of 9-11 year
old children (instrumentalists and non-musicians, not part of the longitudinal study)
corroborates the above structural neuroplastic trends, with the additional finding of
increased grey matter in bilateral occipital lobe and sensorimotor areas in the musically
trained group. The functional tasks yielded more activation of the superior temporal
gyrus (particularly on the right), and the posterior and inferior frontal and middle frontal
gyri (bilaterally), in the 9-11 year old instrumentalists as compared to the 9-11 year old
non-musicians. These functional results corroborate the trends apparent in the 5-7 year
old group. The results of both studies seem to point to the interpretation that much of the
musician structural makeup is experience-induced, though as these studies are not
genetically informative, it must be noted that genetic effects could be involved. The
conclusion that experience is inducing structural changes is supported by literature
showing correlations between structural changes and age of onset of musical lessons
(Elbert et al, 1995, Schlaug et al, 1995, Lee et al, 2003), and/or intensity of practice
(Gaser & Schlaug, 2003, Hutchinson et al, 2003).
In sum, the literature on grey matter differences between musicians and non-
musicians as studied with automated analyses of MRI is full of mixed information. As
discussed briefly at the beginning of this section, it remains unclear as to why there are so
many different results, particularly in such similar populations. In an attempt to shed
20
more light on the information above, results will be briefly reviewed in relation to
instrumentation and gender.
Keyboardists
Male keyboardists show larger areas of grey matter density in bilateral precentral
gyrus (motor cortex), bilateral postcentral gyrus (somatosensory cortex), and premotor
areas (this term is not broken down into neuroanatomical areas; the authors are likely
referring to Brodmann area 6 (BA6), anterior to the precentral gyrus), as well as in
parietal lobe, inferior temporal gyrus, the left cerebellum, left Heschl’s Gyrus, and left
inferior frontal gyrus, when compared to both non-musicians and amateur musicians
(Gaser & Schlaug, 2003).
Both male and female keyboardists show larger areas of grey matter density in
left Rolandic cortex (sensorimotor cortex), bilateral cerebellum, right fusiform gyrus,
right middle orbitofrontal gyrus, left inferior frontal gyrus, left intraparietal sulcus, and
left Heschl’s gyrus when compared to non-musicians (Han et al, 2009) and amateur
keyboardists (James et al, 2013).
Mixed Instrumentation
Male musicians of mixed instrumentation show larger areas of grey matter density
in left inferior frontal gyrus, and an age-related retention of grey matter density in
bilateral dorsolateral prefrontal cortex when compared to non-musicians (Sluming, 2002).
Male and female musicians of mixed instrumentation show larger areas of grey
matter density in right planum temporale and right planum polare when compared to non-
musicians (Bermudez & Zatorre, 2005), increased grey matter density in left anterior
hippocampus (Groussard et al, 2010), and also display a larger overall size of Heschl’s
21
gyrus. Each musician additionally displays a leftward or rightward asymmetry of
Heschl’s gyrus correlated with their pitch perception preference category as explained
above; either a preferences for fundamental or spectral pitch (Schneider et al, 2005).
Child musicians of both genders and of mixed instrumentation show trends in
increased grey matter in sensory and motor areas when compared to non-musicians. This
effect is seen after a few years of private lessons and practice. The children in this study,
like many children of the same age (5-11 years old), are taking private lessons and not yet
participating in ensembles. These children also seem to be performing slightly better on
aptitude tests than their matched controls (Schlaug et al, 2005, Schlaug, 2006, Hyde et al,
2009).
3. White Matter Volume (Diffusion Tensor Imaging (DTI))
To date, only six studies have looked specifically at white matter volume differences
between musicians and non-musicians; one further study (Loui et al, 2010) has
investigated white matter volume differences between musicians with absolute pitch and
musicians without absolute pitch. The studies comparing musicians and non-musicians,
each with slightly different subject populations, have shown results that, much like the
above grey matter studies, do not always agree.
Schmithorst and Wilke published a study in 2002 in which they investigated white
matter differences (accounted for by differences in fractional anisotropy (FA)) in a small
group of musicians (n=5) and non-musicians (n=6). FA increases were reported in the
genu of the corpus callosum and in the cerebellum in musicians as compared to non-
22
musicians. Lower FA values were seen in corona radiata and the internal capsule in
musicians, compared to non-musicians. The authors attribute the areas of increase and
decrease to the continuous repetition of fine motor movements required for musical
performance and the automated nature of those movements after years of practice.
Another DTI study, published in 2005 by Bengtsson and colleagues, found both
increases and decreases in FA in varying regions for 8 male pianists when compared to 8
male non-musicians. The areas found to have changes in FA varied in relation to the time
of life during which each pianist practiced the most (childhood, adolescence, or
adulthood). Higher levels of practicing during childhood correlated with increased FA in
the right posterior internal capsule and in the isthmus and body of the corpus callosum.
Higher levels of practicing during adolescence correlated with increased FA in the
splenium and decreased FA in the body of the corpus callosum. Higher levels of
practicing during adulthood correlated with increased FA in arcuate fasciculus and
decreased FA in the anterior limb of the internal capsule. The authors propose that the
largest amount of areas affected occurred when the pianists practiced more during
childhood, because the practicing occurred during a time when white matter fiber tracts
are maturing, and thus most susceptible to change induced by experience. That is not to
say that experiences as an adult do not have the ability to alter white matter
microstructure; rather, fewer regions are affected during adulthood due to the fact that
white matter tracts have already reached maturity.
Han’s (et al, 2009) study mentioned before on keyboardists (n=19), both male and
female, and gender and age-matched non-musicians (n=21), showed, as mentioned,
higher FA in the right posterior limb of the internal capsule (PLIC), the midbrain area of
23
the brain stem, and the left inferior frontal gyrus. However, there were no areas of lower
FA in musicians than in non-musicians.
Imfeld (et al, 2009) investigated white matter differences specifically in the
corticospinal tract of musicians of mixed instrumentation, both with (n=13) and without
(n=13) absolute pitch, and non-musicians (n=13). Subjects of both genders were
included. The authors report a bilateral decrease in FA in the cerebrospinal tract in
musicians compared to non-musicians, and a general rightward asymmetry in the
corticospinal tract for all groups. Additionally, when the musician group was divided by
age of onset of musical training (pre-7 years old vs. 7-10 years old), increased diffusivity
of the corticospinal tract was found for earlier onset musicians. No effect of absolute
pitch was found. FA findings were confirmed using voxelwise, ROI, and slicewise
analyses; for the latter approach, FA values for the corticospinal tract were calculated
separately for each hemisphere, within each transverse slice of the brain; this shows
variation within the entire tract, and not just within individual ROIs, Imfeld’s
multifaceted statistical approach strengthened these somewhat unexpected results, which
fall in line with those of Schmithorst, but not those of Bengtsson or Han, reported above.
Abdul-Kareem (et al, 2011) compared the cerebellums of musicians (n=10;
keyboard and/or strings) and non-musicians (n=10), using both FA values and
streamlines, which are generated via streamline tractography, a technique meant to help
discriminate between crossing fibers within voxels. Both genders were included. Results
revealed increases in white matter volumes in the right superior and middle cerebellar
peduncles, and a total white matter volume increase in right cerebellum for musicians as
compared to non-musicians. A separate automated grey matter analysis (performed using
24
Freesurfer) revealed no significant differences in grey matter density in the cerebellum
between the groups.
Finally, Steele (et al, 2013) investigated corpus callosum FA differences in early-
trained musicians (ET; n=18), late-trained musicians (LT; n=18), and non-musicians
(n=17). ET musicians began their musical training prior to age 7; LT musicians began
their musical training after age 7. All subjects were scanned in adulthood, and both
genders were included. Instrumentation was not reported, although the authors state that
all musicians in the sample were currently playing instruments (possibly more than one)
that required bimanual coordination. The authors found that ET musicians showed
significantly higher FA in the posterior midbody and anterior isthmus of the corpus
callosum when compared to LT musicians. These FA values correlated negatively with
age of onset of musical training; that is, the earlier the musician began training, the higher
the FA values in adulthood. Interestingly, there was no difference between the LT
musician group and the nonmusician group, indicating that there might be a critical
period for development of this area of the corpus callosum that closes around age 7. As
discussed in the paper, this region connects the sensorimotor cortices of each hemisphere,
and is undergoing great development related changes between ages 6-8. Additionally, the
authors reported that temporal lobe FA values correlated with age of onset of musical
training in both musician groups. As Steele discusses, fibers in this region include those
that connect auditory cortices to motor and parietal cortices via the arcuate fasciculus.
Taken together, the authors make a strong case for a critical period for the development
of fiber tracts that support the type of sensory and motor integration so important to
professional musicians (Steele et al, 2013).
25
Findings from white matter volume (as estimated by FA) studies on musicians
and non-musicians illuminate possible methodological limitations. The difficulty in
interpretation posed by crossing fibers continues to be a problem, despite technical
improvements. The corticospinal tract extensions in hand and face areas, for instance, lie
in regions with an enormous amount of fiber crossings. Newer techniques such as two-
tensor streamline tractography (Qazi et al, 2009, used in Abdul-Kareem et al, 2011),
promise to help resolve this problem. At the moment, we find corroboration between at
least two of the above three studies for increased FA in the right posterior internal
capsule, the right cerebellum, and both increased and decreased FA in various portions of
the corpus callosum (appearing to be mostly due to age at which the musicians practiced
the most). Finding increased FA values in right posterior IC, the site of the descending
motor fibers (corticospinal tract) from the right motor region towards the brainstem, in
musicians, can be explained by the fact that pianists spend much time developing
dexterity in the left hand. An absence of difference between musicians and non-musicians
in the left posterior internal capsule could be explained by the commonality of base use
of the right hand in the two groups. However, a bilateral decrease in cerebrospinal tract
FA in musicians compared to non-musicians (as reported in Imfeld et al, 2009) certainly
is difficult to explain. Schmithorst and Wilke suggest that decreases in the internal
capsule be interpreted as a result of motor entrainment; that the automation of fine motor
movements required for musical performance after years of practice has decreased the
need for resources, and thus decreased FA in the cerebrospinal tracts. This explanation
does not accord with others in the field of experience-dependent structural
neuroplasticity; rather it is more likely that these results in FA are indicative of some
26
other white matter microstructure changes in response to musical experience over time.
Regarding the corpus callosum, the increased FA in the genu (Schmithorst & Wilke,
2002) of musicians compared to non-musicians could be explained by the fact that the
white matter fibers traveling through that region subserve orbitofrontal and anterior
frontal areas, many of which have been implicated in different aspects of musical
processing (Sarnthein et al, 1997, Halpern & Zatorre, 1999, Sakai et al, 1999, partially
reviewed in Schmithorst & Wilke, 2002). Bengtsson’s findings of increased FA in the
splenium and decreased FA in the body of the corpus callosum associated with high
levels of practice during both childhood and adolescence are interpreted as a result of
development-related changes in the white matter tracts of these frontal areas, which
connect to auditory and visual processing regions (splenium), as well as the sensorimotor
cortices (body). Bengtsson suggests that, at least in terms of the corpus callosum body, a
possible increase in callosal axon diameter might yield a decrease in MRI signal intensity
in this area (Bengtsson et al, 2005). Of course, if Steele’s (et al, 2013) critical period
holds, it is possible that FA values in the body of the corpus callosum have more to do
with the age of onset of musical training than with how many hours one practiced
throughout childhood and adolescence. Ultimately many of the above results are
conflicting and merit additional exploration, particularly in the cerebrospinal tract and
corpus callosum, using diffusion MRI.
4. Summary
To conclude this review of neuroanatomical studies performed on musicians and non-
musicians, the following observations can be made regarding brain areas implicated
27
across different techniques (or, in the case of somatosensory cortex, at least across
several studies within the same technique):
Corpus Callosum
The anterior corpus callosum is larger in musicians when compared to non-
musicians (Schlaug et al, 1995, Lee et al, 2003) and this finding is supported, at least in
male musicians, by increased FA in the genu of the corpus callosum of musicians
compared to non-musicians (Schmithorst & Wilke, 2002).
Motor Cortex (Precentral Gyrus)
The two measurements employed in the volumetric literature - intrasulcal length
of the precentral gyrus (Amunts et al, 1997) and hand knob (Bangert, 2006) – both show
larger portions of the primary motor cortex in male musicians than in non-musicians.
Additionally, choice of instrument affects the laterality of these larger volumes,
particularly in the hand knob region; pianists revealed a leftward asymmetry of the hand
knob, while strings players revealed a rightward asymmetry of the hand knob. These
primary motor cortex results are supported by both VBM and DTI studies. Male
professional pianists show larger grey matter density in the primary motor cortex than
male amateur musicians (pianists) and non-musicians (Gaser & Schlaug, 2003). This
result is additionally strengthened by an increase in grey matter density in sensorimotor
areas of children taking music lessons as compared to same age children not taking music
lessons (Schlaug et al, 2005 & Schlaug, 2006). Finally, two DTI studies found increased
FA in the right posterior internal capsule in pianists, in accordance with the motor cortex
results (Bengtsson et al, 2005, and Han et al, 2009). It must also be noted that two further
28
DTI studies reported decreased FA bilaterally in the internal capsule and cerebrospinal
tract (Schmithorst and Wilke, 2002, Imfeld et al 2009).
Somatosensory Cortex (Postcentral Gyrus)
The grey matter density in somatosensory cortex is larger in musicians than in
non-musicians (Gaser & Schlaug, 2003, Han et al, 2009 (adults), and Schlaug et al, 2005
and 2006, Hyde et al 2009 (children)).
Cerebellum
Hutchinson’s group found larger absolute and relative cerebellar sizes in male
keyboardists, but not in female keyboardists (Hutchinson et al 2003). This is supported
by the finding of larger grey matter in the right cerebellum of male and female
keyboardists compared to (male and female) non-musicians (Han et al, 2009), the finding
of larger grey matter in the left cerebellum of male professional keyboardists compared to
both male amateur musicians (keyboardists) and non-musicians (Gaser & Schlaug, 2003),
the finding of bilateral posterior cerebellum grey matter in professional keyboardists
compared to amateur keyboardists and non-musicians (James et al, 2013), and the finding
of increased total white matter in the right cerebellum of male and female keyboard and
strings players (Abdul-Kareem et al, 2011). There is a suggestion of a gender effect in
these studies, but it is difficult to be sure about it without more detailed investigation.
Heschl’s Gyrus
Schneider’s group found a correlation between the absolute size and grey matter
volume of Heschl’s gyrus and musical ability in musicians of both genders as compared
to non-musicians. There also was an asymmetry in lateral Heschl’s gyrus that correlated
with pitch perception preference (Schneider et al, 2005). This is additionally supported by
29
the finding of greater grey matter density in Heschl’s gyrus bilaterally for male
professional pianists compared to male amateur musicians (pianists) and non-musicians
(Gaser & Schlaug, 2003), and the finding of great grey matter density in left Heschl’s
gyrus correlating with level of expertise in male and female pianists, both professional
and amateur, compared to male and female non-musicians (James et al, 2013).
Inferior Frontal Gyrus
Sluming’s group’s found larger grey matter volume in the left pars opercularis
and larger grey matter density in the left inferior frontal gyrus of male musicians
compared to non-musicians (Sluming et al, 2002, Abdul-Kareem et al, 2011). Gaser &
Schlaug on the other hand found increased grey matter density bilaterally in the inferior
frontal gyrus in male pianists compared to non-musicians (Gaser & Schlaug, 2003), and
James et al reported increased grey matter density in the left inferior frontal gyrus of
male and female pianists, both professional and amateur, compared to male and female
non-musicians, and these differences correlated with level of expertise (James et al,
2013). Schlaug’s longitudinal and correlational studies in children show increased
functional activity in the inferior frontal gyrus of male and female musicians in response
to musical tasks, when compared to non-musicians. Although these are functional results,
structural changes in areas undergoing functional changes in the same cohort of children
have also been reported (Schlaug et al, 2005, 2006, Hyde et al, 2009). Finally, these
results are supported by Han’s study, where increased FA in the left inferior frontal gyrus
of both male and female pianists was found when compared to non-musicians (Han et al,
2009).
30
While many more areas of the brain were discussed as being larger in musicians
compared to non-musicians in the studies reviewed, the above six regions (corpus
callosum, primary motor cortex, somatosensory cortex, cerebellum, Heschl’s gyrus, and
inferior frontal gyrus) were the only regions to achieve significance in several of the
studies and across several different methods of analysis. As discussed earlier, there is a
need to perform neuroanatomical studies on both male and female musicians, and on
many different types of instrumentalists. Though there seems to be a trend for change, the
extant literature mostly addresses male musicians who play one or two types of
instruments (keyboard and strings). Studies should address both genders and a larger
variety of highly skilled musicians (vocalists, composers, and conductors included) in
order to draw generalizable conclusions about how musicians’ brains may differ from
non-musicians’ brains.
ii. Influence of Genetics
In relation to the structural differences analyzed above, one question continues to arise:
do these anatomical differences between musicians and non-musicians exist because of
genetic influence, or is intense training during a critical period of cognitive development
to blame? In other words, are individuals with a genetic predisposition (and
accompanying predisposed cortical organization) toward musical proficiency both drawn
to musical training and inspired to continue with disciplined practice? Or are the
structural differences so far reported due specifically to musical training? There is
evidence that structural changes in musician’s brains are strongly correlated with age of
31
onset of training (Elbert et al, 1995, Schlaug et al, 1995, Lee et al, 2003), and intensity of
practice (Gaser & Schlaug, 2003, Hutchinson et al, 2003), which at least points towards a
strong environmental influence, and possibly a gene-environment correlation or
interaction.
Arguably the best way to answer the question of nature vs. nurture is through
genetically informative longitudinal studies with children. Recent results from Schlaug’s
longitudinal study (reviewed in the previous section) indicate that there is perhaps more E
than G influence in this Gene-Environment interplay. Children in his study taking
musical lessons were matched to children not taking musical lessons on three factors:
age, socioeconomic status, and verbal IQ. Schlaug included extensive testing before
lessons began, including cognitive, musical, and motor abilities. Anatomical analysis was
also performed prior to the beginning of lessons, to investigate pre-existing
neuroanatomical differences between the groups. None were found (Schlaug et al, 2005,
Schlaug 2006). Thus, the gains in aptitude and minor neuroanatomical changes exhibited
in the children taking musical lessons, but not in the control group, suggest that the
environment plays a significant role in shaping the musician’s brain. That said, the
musical backgrounds of the children’s families were not taken into consideration, making
it more difficult to draw conclusions about the potential role of genetics. It remains to be
seen if children continue to exhibit neuroanatomical and functional changes after a
possible instrument switch or stopping musical lessons altogether, as well as if those
children that go on to pursue lifelong musical careers are those from families with strong
histories of musical proficiency.
32
It is commonly accepted that music tends to run in families. Anecdotal evidence
exists everywhere; one only has to look to past (and present) celebrity “musical families”
like the Bach, MacCrimmon, Mozart, Strauss, and Mehta (and perhaps vonTrapp)
families. As discussed in Ukkola-Vuoti (et al, 2013), several studies have recently
revealed genetic components to both musical ability and absolute pitch, (AP, known as
“perfect pitch” to musicians), the ability to correctly identify any pitch without a
reference point (e.g., musical context or starting note). Some evidence exists to support
the idea that absolute pitch results from a combination of genetic predisposition and
environmental exposure during critical periods of development (Zatorre, 2003, Trainor,
2005), and for instance, in relation to tonal languages (Deutsch, 2006). The high
incidence of AP in individuals brought up in the environment of tonal languages
compared to the low incidence of AP in most relative pitch languages supports Steven
Mithen’s hypotheses regarding the origins of music and language. Mithen proposes that
an evolutionarily older system of communication comprised of music-like vocalizations,
which he calls “Hmmmmm (Holistic, manipulative, multi-modal, musical, mimetic),”
was replaced in homo sapiens by language. Mithen’s idea of Hmmmmm has
Neanderthals and earlier homo species singing and grunting their way through life. The
development of largely relative pitch-based modern languages, Mithen argues, would
have decreased the natural propensity toward viewing all auditory stimuli as musical,
thus eventually enabling humans to use music for more than simply species survival
reasons. The development of relative pitch-based languages would have also naturally
decreased the rate of individuals with AP, as perfect pitch became less necessary for
33
communication (Mithen, 2006).
1
The concept that this aspect of musical ability might be
more universally achievable than previously thought (with exposure to some form of
tonal language early in life) concurs with the idea that humans might be hard-wired for
musical competency, at least in musical understanding. The idea of a universal musical
competency had been suggested before (Cross, 1999). This concept is supported, at least
in theory, by the variety of evolutionary purposes proposed for music: social bonding
(Cross, 1999, Brown, 2000), infant-mother bonding (Dissanayake, 2000), memory
facilitation (Bharucha et al, 2006), and, primarily in the melodies of ‘motherese’ and
infant-directed speech and singing, the priming processes of humans for linguistic
acquisition (Koelsch & Siebel, 2005). As Huron notes, “there is no human culture known
in modern times that did not, or does not, engage in recognizably musical activities”
(Huron, 2001). All this evidence seems to suggest that in fact the propensity to achieve
AP may be genetic and that all it needs is the right environmental demands to manifest
itself. All of it also points toward a possible universal human propensity for competency
in musical understanding, if not musical ability.
2
Whether or not this universal
competency in musical understanding translates into a shared propensity toward
instrumental proficiency remains to be seen. Schlaug’s study, at least, seems to point in
this direction. The fact that his cohort of children (having no pre-existing cognitive,
musical, motor, or neuroanatomical differences to demarcate them from the control
group) do evolve to exhibit both cognitive and neuroanatomical changes directly
1
Portion of paragraph beginning with “Mithen proposes that…” is both borrowed and
paraphrased from M. Miles’ answer to qualifying examination question #2, December,
2006.
2
Portion of paragraph beginning with “The idea of a universal…” is both borrowed and
paraphrased from M. Miles’ answer to qualifying examination question #3, December,
2006.
34
correlated with their musical training, presents a strong argument, if not for a universal
musical competency, then certainly for the large role that the environment plays in the
shaping of the human brain, particularly in relation to music. Such an argument would be
further addressed by studies involving adult novice musicians, following roughly the
same protocol as the one used for the children learning an instrument; it would help to
determine the extent to which critical periods of development matter for altering the
structure of a musician’s brain.
iii. Non-Musical Experience-Dependent Structural Plasticity
Before delving into this field, it is important to briefly note that several studies have
suggested that grey matter volume naturally decreases with age, while white matter
volume often increases up to the mid-50s (reviewed in Allen et al, 2005). Studies on
‘experience-dependent structural plasticity’ (Aydin et al, 2007), however, reveal
structural neuroplasticity (in terms of increasing grey matter) in adults of varying ages,
and in a variety of regions. These studies include a wide variety of learned skills, thus
demonstrating that many regions of the brain seem capable of structural neuroplasticity in
adulthood. For purposes of this review, the studies will be grouped according to general
skill-set.
1. Learning and Memory
Maguire’s group released a now quite famous study in 2000 on London taxi
drivers, generating a lot of debate regarding experience-dependent structural plasticity.
35
Sixteen right-handed male London taxi drivers and fifty right-handed non-taxi driver
controls underwent structural MRI scanning. London taxi drivers were specifically
chosen because of the extensive navigational training each undergoes to gain what is
known as “The Knowledge” (a profound knowledge of thousands of London locations
and landmarks). The learning typically takes approximately two years, after which an
aspiring London taxi driver can take exams to obtain a license. VBM analyses of the
subjects yielded significantly larger posterior hippocampal grey matter volume in taxi
drivers vs. controls; the size of this volume positively correlated with the amount of time
spent as a taxi driver. Controls exhibited larger volume in the anterior hippocampus, also
bilaterally, than the taxi drivers. Animal studies have implicated the posterior
hippocampus as important for spatial navigation (reviewed in Maguire et al, 2000). The
authors posit that being a London taxi driver for a period of time affects a change in
distribution of grey matter volume in the hippocampus (Maguire et al, 2000).
After performing the above study, Maguire’s group investigated whether or not
the hippocampal differences revealed in the study were due to being on “The
Knowledge” and its constant practical application, or to having an excellent innate sense
of direction and navigational skill (as no doubt possessed by many Londoners, and which
may predispose a person to becoming a taxi driver). The second study performed by the
Maguire group thus placed 26 right-handed non-taxi-driving males with varying levels of
navigational expertise in a virtual environment for navigation. The 26 subjects were
investigated for structural differences before and after utilizing the environment. The aim
was to document a positive correlation between hippocampal grey matter volume and
navigational skill, as determined by tests regarding and utilizing the virtual environment.
36
The finding of a lack of significant hippocampal grey matter differences correlated with
navigational skill in the group (after utilizing the virtual environment) suggests that it is
the process (the incredibly large amount of learning required to gain “The Knowledge”)
of learning and utilizing internal spatial maps every day that induces structural change,
and not an excellent innate sense of direction (Maguire et al, 2003).
Aydin and colleagues studied, using VBM, a group of academic mathematicians
(n=26) and of other academics (n=23) because the inferior frontal and inferior parietal
lobules have been shown to be involved in arithmetic processing and other types of
mathematical thinking (Aydin et al, 2007). They found larger grey matter density in left
inferior frontal and bilateral inferior parietal lobules of academic mathematicians when
compared to other academics. The difference in amount of grey matter in the right
inferior parietal lobule was positively correlated with years spent as an academic
mathematician. (Aydin et al, 2007).
A 2004 study by Mechelli and colleagues with English-Italian and Italian-English
bilinguals revealed an increase in grey matter density in the left inferior parietal cortex in
bilinguals when compared to monolinguals. This increase was both correlated with age of
acquisition of the second language and proficiency in that second language. In the same
study the inferior parietal area was also shown to be activated during fMRI studies of a
verbal fluency task (Mechelli et al, 2004).
Yet another study related to experience-dependent structural plasticity was
published in 2006 by Draganski and colleagues. This study involved medical students
(n=38) undergoing the intense process of studying for their medical examinations. They
were scanned at three separate time points: 3 months prior to their exams, 1 or 2 days
37
after their exams, and 3 months after their exams. During the study prior to the exams,
grey matter increases were detected (within subject comparison and also matched
controls (n=12)) in bilateral posterior and lateral parietal cortex, and bilateral
hippocampus. The authors discussed the parietal increases in light of suggestions that
these areas are associated with information transfer into long-term memory, storage of
visual short-term memory, and memory retrieval (Draganski et al, 2006). An additional
interesting finding was that the grey matter changes in the hippocampus actually had
increased further at 3 months post exams. The authors interpret this increase as possibly
due to some form of neurogenesis in the hippocampus, which would likely cause a
delayed increase in detectable grey matter at the macroscopic level (Draganski et al,
2006). Of course, such an interpretation remains purely theoretical, as it would be
impossible to investigate microscopic brain changes with the tools currently available for
the study of living subjects. Furthermore, it is important to point out that the findings
from the final scan, taken months after the exams, may not mean that the brain continued
to change without any further learning activity, as medical students continue the learning
process at a high rate. In other words, despite entering the semester break post task, it is
unlikely that these students truly ceased the process of learning new information in
between scans. It would probably be beneficial to follow this study up with high school
juniors or seniors studying for advanced placement exams; the months-later scan could be
performed in the summertime, when school is not normally in session.
2. Motor Skills
Draganski’s group published an earlier study in 2004 on adults (n-12) learning the
heavily skilled task of juggling over a period of 3 months. Subjects were scanned a total
38
of 3 times: before starting to learn how to juggle, after becoming skilled at it (at the end
of 3 months of training), and after a 3 month cessation of all juggling activities. Increases
in grey matter (using VBM) were seen in the jugglers compared to controls (n=12) at the
end of 3 months of training bilaterally in the middle temporal area of the visual cortex
(V5; the visual motor processing area) and the left posterior intraparietal sulcus. These
increases were directly correlated with juggling performance, and decreased after 3
months of no juggling (Draganski et al, 2004). Two further investigations into juggling
training have corroborated the grey matter increases in V5, and subsequent decreases
following the cessation of training (Boyke et al, 2008, Driemeyer et al, 2008). One
additional study combining both VBM and DTI analyses found FA increases in the
juggling group (compared to controls) in the right posterior intraparietal sulcus, and grey
matter density increases in the juggling group in medial occipital and parietal lobe areas
overlying the white matter increases. These grey matter and white matter changes
remained after a four-week period of no juggling (Scholz et al, 2009).
Wei, Zhang, Jiang, and Luo published a study in 2011 demonstrating increased
cortical thickness in the left superior temporal sulcus, the right orbitofrontal cortex, and
the right parahippocampal gyrus in professional divers when compared to non-athletes.
The findings in parahippocampal gyrus were positively correlated with training
experience (Wei et al, 2011).
Jäncke and colleagues found grey matter increases in the brains of golfers in
premotor and parietal areas, and lower white matter volume and FA in the corticospinal
tracts, when compared to non-golfer controls. Furthermore, these effects were modulated
by intensity of experience (amateur vs. professional status) (Jäncke et al, 2009).
39
One further study demonstrated decreases in white matter in the corticospinal
tract in relation to experience. Using both VBM and DTI methods, Hänggi additionally
found lower grey matter in the left premotor cortex, supplementary motor area, putamen,
and superior frontal gyrus in the brains of professional ballet dancers compared to
controls. White matter decreases were demonstrated not only in the corticospinal tract,
but also in the internal capsule, corpus callosum, and left anterior cingulate. FA decreases
were found in bilateral premotor cortex. As with many of the above studies, age of onset
of dance training modulated the effect. Though these decreases were unexpected, based
on the extant literature, the authors suggest several possible ways to interpret them,
including the possibility of an excess in fiber crossings, which can negatively affect FA
values (Hänggi et al, 2010).
3. Repetitive Magnetic Stimulation (rTMS)
One study using repetitive transcranial magnetic stimulation (rTMS) on the
superior temporal gyrus of healthy people (n=18) revealed grey matter increases
(measured with VBM) in the contralateral auditory cortex in as little as 5 days after the
start of daily rTMS intervention. Although this is not skill-dependent, the results do
indicate the rapid ability of the adult brain to adjust its own structure as a result of
experience (May et al, 2007).
4. Summary
Although the Maguire, Aydin, Mechelli, and Wei studies included some subjects
who acquired their skills in childhood, the fact that the effect size correlated positively
with the amount of time spent honing those skills indicates that some structural
neuroplasticity is occurring in adulthood, and in areas specifically relevant to the skills in
40
question. Taken together, the above studies suggest that structural neuroplasticity,
reflected as increased grey matter density, is not only possible in adults, but it is
achievable over a relatively short time span and sometimes maintained for a period of
time after task completion. The task does not necessarily have to be novel (as for instance
learning to juggle); as with the medical students studying for their exams, if the task is a
familiar one, but more frequent and intense than in prior periods, structural
neuroplasticity may still be found. Additionally, the fact that experience-dependent
structural plasticity can be observed in relation to both motor tasks as well as learning
and memory tasks bodes well for an investigation into potential neuroanatomical changes
in collegiate musician brains caused by conservatory training.
iv. Possible Microscopic Mechanisms
1. Histology
It stands to reason that if a person, through some type of activity, intensifies the function
of one part of his/her brain over the rest of the same brain on a regular basis for an
extended period of time, that the part of the brain in continuous use (above and beyond
what was previously normal function for said person), will respond to said use with a
variety of structural changes to accommodate the functional demand in a more efficient
manner. Such argument is supported both by the correlation of structural differences in
musicians and non-musicians with the age of onset of musical training (and years of
continuous practice), as well as by the growing body of literature devoted to other kinds
of experience-dependent structural plasticity. While some of the structural differences
41
referred to in the earlier section may be due to, or enhanced by, the fact that the brains
were still very much in development when music lessons were started, the structural
differences referred to in the latter section had occurred almost exclusively in adults well
beyond the normal stages of neurodevelopment. The question of what, exactly, at the
microscopic level, is the basis for these macroscopic structural changes, remains to be
answered. One recent review has attempted quite thoroughly to elucidate candidate
mechanisms for both grey matter and white matter changes observed in the experience-
dependent structural neuroplasticity literature (Zatorre et al, 2012); these mechanisms
will be reviewed here. Limitations of current technology force one to pursue this question
by reviewing related animal research.
2. Neurogenesis
Kempermann and colleagues have performed several studies utilizing adult mice
to show that enriched environments can induce the growth of new neurons, particularly in
the dentate gyrus of the hippocampus and in the olfactory bulb. Mice in environments
that promote learning and social interaction were compared to mice in standard
laboratory cages (equipped with food, water, and few other mice). This experiment took
place over both short and long time periods, and revealed that, while short-term exposure
to the enriched environment produces marked neurogenesis, long-term exposure to the
same type of enriched environment produces adult hippocampal neurogenesis at a rate
five times higher than in controls. The mice in enriched environments also showed a
decrease in age-related degeneration in the dentate gyrus when compared to controls
(Kempermann et al, 1997, Kempermann et al, 1998, Kempermann et al, 2002,
Kempermann et al, 2004, Song et al, 2005, Garthe & Kempermann, 2013).
42
Zatorre et al argues that, as neurogenesis is only known to occur regularly in the
hippocampus, it is unlikely to be the major molecular component in changes in grey
matter seen in neuroplasticity studies, at least outside of the hippocampus. The authors
suggest that changes are more likely due to gliogenesis, synaptogenesis, or experience-
related changes in vasculature (Zatorre et al, 2012).
3. Synaptogenesis
Three studies in adult rats have revealed synaptic plasticity in direct response to
motor training. Black et al showed in 1990 that acrobatic training produced
synaptogenesis in the paramedian lobule of the cerebellar cortex, while normal physical
exercise and inactivity did not. Interestingly, the active rat group participating only in
physical exercise (not acrobatic training) developed a greater density of blood vessels
(angiogenesis) in the same region, suggesting that increased normal activity without new
learning elicits angiogenesis (Black et al, 1990).
Jones and colleagues used an acrobatic task to produce synaptogenesis in the
forelimb sensorimotor cortex of adult rats on the side opposite of a unilateral lesion
created by the experimenters (Jones et al, 1999). Kleim et al trained adult rats on two
reaching conditions (skilled and unskilled). The brains of the rats performing the skilled
reaching condition revealed synaptogenesis followed by larger distal forelimb
representations in motor cortex after 10 days of training, indicating that both synapse
formation and the reorganization of motor maps can occur in a relatively short time in
response to learned motor skills (Kleim et al, 2004).
Trachtenberg et al trimmed adult mouse whiskers in a checkerboard pattern,
changing each mouse’s interaction with their environment. The trimmed-whisker mice
43
showed increased synaptic turnover in the barrel cortex as compared to normal-whisker
mouse controls (Trachtenberg et al, 2002).
4. Axonal Sprouting
Moving beyond the realm of rodent mammals, Hihara and colleagues trained
adult monkeys on a tool-use task (using a rake to retrieve food) for six weeks. The brains
of the monkeys trained in tool-use revealed the emergence of cortical afferents from the
temporo-parietal junction and ventrolateral prefrontal areas to the anterior bank of the
intraparietal sulcus. The brains of the control monkeys revealed no significant changes
(Hihara et al, 2006).
Zatorre et al also suggest possible mechanisms of axonal pruning or re-routing.
Evidence for both exists in the extant literature, and any of these three axonal
mechanisms could affect changes in white matter (Zatorre et al, 2012).
5. Other
Several other microscopic processes that could be underlying macroscopically
measurable differences in anatomical structure in response to experience have been
suggested: the proliferation of microglia, the swelling of neurons, and/or an increase in
capillary density (in accordance with the angiogenesis found in Black et al, 1990), and
possible changes in myelination throughout adulthood (Terrazas and McNaughton, 2000,
Aydin et al, 2005, Schlaug 2006, Zatorre et al, 2012).
Notably, microstructural changes in adult animals in response to experience have
been reported in many regions homologous to those found to have macrostructural
change in adult humans in response to experience: hippocampus, motor cortex,
44
somatosensory cortex, and cerebellum. It remains possible that different processes are at
work in different regions of the human brain; neurogenesis may be increasing grey matter
volume in the hippocampus, for example, while a similar increase in somatosensory
cortex might be due to synaptogenesis. Without advancement in technology for human
research – or a supply of musically-gifted primates – it is difficult to conclude what
exactly is occurring histologically in humans in direct response to experience. The most
important message to extract from the animal experience-dependent plasticity literature is
that microstructural changes can indeed occur in adulthood, through several different
mechanisms.
v. Summary
Reliable neuroanatomical differences in professional musicians – compared to non-
musicians – at the adult level, some of which can be corroborated by results in children
taking music lessons, have been found in motor and auditory related areas as well as the
inferior frontal gyrus. The extant literature suggests that while it remains unclear exactly
what histological mechanisms are driving the development of these differences, it is
likely that the experience of musical training itself, and not simply a genetic
predisposition toward musicianship, is both triggering and supporting these plastic
changes in the musician’s brain. The gap between childhood music lessons and
professional musicianship remains to be addressed, and must be addressed, in order to
achieve a more complete understanding of when and why these differences appear. The
45
following chapters describe a study undertaken to help address this gap in the field of
musical experience-dependent structural neuroplasticity.
46
CHAPTER 2
A. INTRODUCTION
Many of the studies reviewed in the previous chapter (both human and animal)
suggest that various types of experience can alter brain function in a manner that will
eventually become observable at a gross structural level. Returning to the type of
experience discussed at the beginning of that review, the unique experience of becoming
and remaining a musician (musical training, practice, and performance in a variety of
settings) has been shown to elicit neuroplastic changes in the corpus callosum, primary
motor cortex, somatosensory cortex, cerebellum, Heschl’s gyrus, and the inferior frontal
gyrus. Although studies have been performed on both adults and children, the studies
have been focused primarily on male musicians, and often only keyboard and/or strings
players. Additionally, the age range of the adults studied has been quite wide. Little work
has been done on musicians in the intensive learning process that comprises professional
musical training. While it is true that professional post-college musicians keep their skills
honed through lessons, summer programs, and continued practice and performance, it is
during their time in conservatory that they often relearn how to think about music;
intensive theoretical, aural, and historical training is added atop what is usually a long
personal history of practice and performance for each musician. The study described here
seeks to answer the question: Does the experience of four years of conservatory training
induce observable structural changes?
As reviewed before, the differences between musician and nonmusician brains
show positive correlation with the age of onset of musical training. It is likely that
observable motor area differences begin occurring in childhood, while differences in
areas such as the inferior frontal gyrus may be revealed only after musicians have
47
undergone proper theoretical training. This hypothesis is based on evidence showing the
involvement of the inferior frontal gyrus in functional processing of the syntax and
semantics of music (Sergent et al, 1992, Platel et al, 1997, Maess et al, 2001, Parsons,
2001, Koelsch et al, 2005). Learning how to process syntax and semantics of music at a
professional musician level requires more than just lessons and practice sessions alone.
Some musicians do undergo theoretical training prior to entering conservatory schooling.
Indeed, many take Advanced Placement music theory courses and examinations at the
high school level. While such activity may induce structural changes in pre-college
students, it seems likely that the combined forces of extensive practicing (often up to
eight hours daily), performing, and theoretical, aural, and historical coursework are what
change the way musicians think about music, thus probably inducing an increased
amount of functional activity in areas such as the inferior frontal gyrus. This functional
activity at sustained increased levels (as compared to pre-college levels) may lead to
structural changes in grey matter density and/or white matter volume in parallel with the
increased functional usage of this brain region. Does such pre-academic training
introduce measurable structural changes in music students compared to students of a non-
music field as they enter college; and is such a difference amplified at the end of college?
In music majors at the conservatory level, one would expect to see structural
neuroplasticity that relates linearly to years of personal experience. In Schlaug’s
longitudinal study discussed before, four years of childhood musical training were
enough to produce general increases in grey matter as well as significant functional
processing increases, measured through musical tasks performed during fMRI scanning.
In post-recital senior conservatory musicians at or around the time of graduation
48
(compared to entering freshmen conservatory musicians), I expect to see structural
changes in motor areas (grey matter increases in the motor cortex (laterality dependent on
instrument of choice), bilateral cerebellum, bilateral somatosensory cortex, and white
matter increases in the corpus callosum (likely anterior) and posterior internal capsule
(again, laterality dependent on instrument of choice) when compared to freshman
musicians; furthermore, I expect a correlation not just with conservatory training but also
with total years of experience as a musician, as determined by age of onset of musical
lessons (usually prior to ten years of age); this change should be different from what is
seen in the non-music artistic academic training (namely, training in architecture). In
addition, I expect that the rigors of music school, laid atop a sturdy foundation of years of
practice, will produce increases of grey matter in areas less related to the motor dexterity
needed to play an instrument and more related to the functional processing demands
created by the theoretical, aural, and historical analysis demands of conservatory training:
namely, the inferior frontal gyrus (likely on the left) and bilateral hippocampus. These
changes are expected to be seen in senior music students compared to freshmen music
students and compared to non-music students, although I additionally expect to see
expertise-related increases in 4
th
year vs. 1
st
year architecture majors in bilateral
hippocampal grey matter (due to the great volume of information memorized each
semester), and possible grey matter increases in the bilateral medial orbitofrontal gyrus,
an area linked in one recent study to aesthetic judgment of visually-presented
architectural stimuli (Kirk et al, 2009). Returning to the musicians, grey matter increases
should be seen particularly in the inferior frontal gyrus (IFG) and hippocampus, with a
likely leftward asymmetry. Structural changes in IFG are hypothesized in freshmen
49
music-students compared to freshmen non-music students, as briefly discussed above,
because of the processing shift that occurs when a musician learns large amounts of
theory and music history; an explanation that may be the reason behind the large
differences of grey matter shown in professional orchestral musicians (Sluming et al,
2002). Structural changes in the hippocampus in senior music students as compared to
freshmen music students (and freshmen non-music (architecture) students) are
hypothesized because of the large musical repertoire that must be learned and memorized
during music school, particularly in preparation for a senior recital. It is possible that, due
to the short time span of conservatory laid atop years of private practice, structural
changes in IFG and the hippocampus will be either small in statistical significance or
merely the representation of a trend. I hypothesize structural changes (in both grey and
white matter) in motor related hand areas, as discussed earlier. I expect these changes to
correlate primarily with years of personal experience as a musician. I also expect changes
in bilateral Heschl’s gyrus (HG), in senior music students when compared to freshmen
music students and to senior non-music students, due to intense aural training in
conservatory. I expect the largest changes in grey matter/white matter in HG to occur in
conservatory students with the least amount of pre-conservatory training, given Gaser and
Schlaug’s 2003 finding of larger grey matter density in HG in professional musicians as
compared to amateur musicians (Gaser & Schlaug, 2003). Finally, I expect to find
increased grey matter and white matter in the insula and cingulate cortices of senior
music students as compared to freshmen music students and both cohorts of architecture
majors. The reason to include the insula and cingulate cortices in my hypotheses is their
functional role in emotional processing (Damasio et al, 2000, Phan et al, 2002, Molnar-
50
Szakacs et al, 2006). Emotional interpretation is an aspect of music learning and music
making that is heavily emphasized during conservatory training; performers cannot hope
to convey meaning to their audiences if they themselves have not spent time deciphering
the emotions represented within each musical score.
B. SPECIFIC HYPOTHESES
1. Four years of conservatory training will induce macroscopic structural changes,
both in grey matter and white matter, in university music students.
2. Conservatory musicians will show neuroanatomical differences at a macroscopic
level compared to other university students undergoing similarly rigorous artistic,
but nonmusical, programs of study (architecture). Differences between these two
groups of students will be evident at entry and graduation time.
C. PREDICTIONS
1. To test hypothesis 1, I will compare the brains of 1
st
and 4
th
year university music
students. I predict that there will be increases in grey matter density in the motor
cortex, cerebellum, inferior frontal gyrus, hippocampus, Heschl’s gyrus,
somatosensory cortex, cingulate cortex, and insula, and higher FA values in white
matter in the corpus callosum and posterior internal capsule of 4
th
year vs. 1
st
year
music majors.
2. These structural changes will positively correlate with amount of practice time
(on instrument and time spent performing both aural and written analysis) and
will be modulated by age of onset of musical lessons for each musician.
51
3. To test hypothesis 2, I will compare the brains of each cohort of music majors
with each cohort of architecture majors. I predict that there will be
neuroanatomical differences (greater grey matter density and FA) in musicians,
particularly 4
th
year seniors, as compared to architecture majors in the same areas
listed above in prediction 1.
4. I predict that the differences between music majors and architecture majors will
also be present in the freshmen cohorts, but they will be larger in the senior
cohorts.
5. I predict a difference between keyboard players and strings players in the motor
cortex due to hand usage for both classes of instruments; there will be a rightward
asymmetry in motor cortex for strings players and a relative symmetry in motor
cortex for keyboard players.
D. SUBJECTS
i. Specific Populations
Entering freshmen music majors were matched as closely as possible to senior
music majors on sex, instrument, and performance background. Music majors of both
genders and representing several major instrument types (strings, vocalists, and piano)
were selected in order to help to add information to the current literature on
neuroanatomical differences between musicians and non-musicians.
In order to control for neuroanatomical changes due to university education by
itself, control subjects were selected from the university population of entering freshmen
and fourth year architecture majors. For both architecture and music majors, general
52
education coursework is spread out over the course of the entire degree program, rather
than lumped into the first two years of study. Architecture majors are an ideal control
population because of the similarly rigorous program of study (similar to a conservatory
degree program) they undergo. While music majors spend much of their out of class time
practice of their instrument, architecture majors spend much of their out of class time
practicing architectural skills via rendering programs and model-building. Both degree
programs require spatial skills; musicians must spatially orient themselves aurally in both
solo and group performance and rehearsal settings, both with other musicians and with
the acoustics of a particular performance space. Architecture majors must spatially orient
themselves visually in order to both design and build to-scale models of buildings.
Architecture majors, like music majors, often do not take summers off from their studies;
they improve their craft via internships and summer programs, just as musicians do.
Architecture majors prepare a thesis project prior to graduation, just as music majors
must prepare a recital prior to graduation. Finally, like most arts, both architects and
musicians must adjust the practice of their work to account for constraints of physical
space. While architecture majors work with the constraints of building and design with
real materials in real space, so do music majors work with the constraints of composing
for, and improvising and playing on, real instruments in real space. Thus, architecture
majors in both year 1 and year 4, and with as little prior musical engagement as possible,
were selected as control subjects for this study.
53
ii. Population Characteristics
A statistical power analysis conducted on the extant literature yielded a necessary
subject number of 18 per group. Due to limitations of both time and funding, subject
numbers were limited to 15 per senior group and 10 per freshman group. The architecture
group was further reduced post-recruitment due to a high enough amount of musical
engagement in 8 of the recruited architecture subjects. These 8 subjects (3 freshmen, 5
seniors) were removed from the analysis and put into a middle group for exploratory
analyses alone. This left the control group with7 architecture freshmen and 10
architecture seniors. One additional subject was removed at scan time because of age; the
subject was within the correct age range (17-20 years of age) when recruited, but by scan
time was unfortunately outside the proposed age range. Thus, the final architecture
groups were reduced to 6 architecture freshmen, and 10 architecture seniors. The music
groups remained at their original level: 10 music freshmen and 15 music seniors.
Therefore, the present study has to be viewed as a preliminary investigation into
neuroanatomical differences induced by conservatory training, changing the statistical
expectations to trends rather than statistically significant findings. Both genders are as
evenly represented as possible in each subject group; however, the senior music major
population contains 10 females and only 5 males. Regarding race, it was difficult to
completely limit the inclusion to either subjects of non-Asian descent or, alternatively, to
subjects of Asian descent; such non-mixing of subjects of Asian descent being suggested
due to the likely difference in brain shape of subjects of Asian descent, which may or
may not have an effect on the anatomical analysis of brain structure. The difficulty lay in
the racial demographics of the university from which my subjects were obtained; while
54
the keyboard and strings departments are roughly 50-60% Asian, and the architecture
school is roughly 30-40% Asian, the voice department is roughly 20% Asian. Thus, both
racial groups were included.
Performance background for subjects was assessed via a comprehensive
questionnaire, attached. Two subjects failed to complete the questionnaire (1 architecture
senior and 1 music senior), and one further subject (a music senior) gave very partial
information. Despite repeated attempts to rectify missing questionnaire data, subjects did
not respond to the requests. Descriptive statistical analysis of the remaining
questionnaires was handled in SPSS, and is detailed on tables 1-7.
1. Population Characteristics by Major and Year
Tables 1-2 detail basic population characteristics for architecture freshmen (n=6,
3 women), architecture seniors (n=10, 5 women), music freshmen (n=10, 5 women),
music seniors (n=15, 10 women), and the unintended middle group, architecture students
with significant musical training (AWM), both freshmen (n=3, 1 woman) and seniors
(n=5, 2 women). As can be seen from the tables, age at scan time is roughly the same for
both of the main freshmen groups (18.33 and 18.22 years), and slightly higher in
architecture seniors (22.33 years) as compared to music seniors (21.73 years). The
majority of each group experienced pre-collegiate academic training in their respective
art (music or architecture). While half of the architecture majors have relatives in the
arts, slightly more music majors do so; 8/10 freshmen and 11/15 seniors, respectively. All
music majors who have relatives in the arts have musician relatives, while only two
architecture majors have architect relatives. Age of onset of musical training, years
55
played, and intensity of practice (the average hours per day spent in practice during years
when the instrument was being played) numbers for the primary and secondary
instruments are listed. Statistical comparisons between groups for these factors can be
found in Tables 3-4. Additionally, age of onset of musical training for the first instrument
has been recorded; not all subjects count their first instrument (often piano) as their
primary or even secondary instrument at scan time, but this information paints a clearer
picture of the amount of musical training each subject has received. For example, for the
variable of primary instrument age of onset of musical training, there is no statistically
significant difference between architecture majors and music majors. However, when one
looks at the same group comparison for the variable of first instrument age of onset of
musical training, there is a clear statistically significant difference. As expected, these
music majors have played both their primary and secondary instruments for statistically
significantly more years than the architecture majors (none of whom still play their
primary instrument, and only one of whom still dabbles with their secondary instrument,
as can be seen on Table 5). The music majors (overall and at the freshman level in
particular) have also experienced statistically significantly more intensity of practice on
their primary instrument than the architecture majors. The statistical trend for this
comparison at the senior level is likely due to the higher proportion of voice majors in the
music seniors, as will be discussed later.
56
Table 1: Population Characteristics, by Major and Year
Architecture
Freshmen (n=6)
Seniors (n=10)
Music
Freshmen (n=10)
Seniors (n=15)
Number of women 3 5 5 10
Number of men 3 5 5 5
Age at scan time 18.33 (± 0.52) 22.33 (± 0.71) 18.22 (± 0.44) 21.73 (± 0.88)
Subjects with absolute pitch 0 1 3 0
Age of onset, musical training
First instrument 8.60 (± 2.41) 8.00 (± 1.29) 4.80 (± 0.92) 5.64 (± 1.55)
Primary instrument 9.80 (± 2.86) 9.00 (± 2.94) 8.00 (± 3.30) 9.23 (± 3.59)
Secondary instrument 12.25 (± 2.63) 12.75 (± 2.22) 8.11 (± 4.40) 9.85 (± 4.49)
Years played, musical instrument
Primary instrument 3.80 (± 1.92) 3.00 (± 1.41) 9.30 (± 4.30) 12.62 (± 3.40)
Secondary instrument 3.00 (± 1.41) 3.00 (± 1.15) 9.11 (± 4.20) 8.37 (± 5.77)
Intensity of practice (average hours/day), musical instrument
Primary instrument 0.67 (± 2.89) 1.25 (± 1.06) 1.86 (± 1.29) 1.99 (± 1.12)
Secondary instrument 1.50 (± 0.00) 1.25 (± 1.06) 0.94 (± 0.29) 1.02 (± 0.82)
Subjects with pre-college musical training
Music theory 0 0 7 11
Music history 0 0 1 5
Aural skills (ear training) 0 0 6 8
Subjects with pre-college architectural training 4 7 0 1
Subjects with relatives in the arts 3 5 8 11
Musicians 2 (1 professional) 3 (1 professional) 8 (3 professional) 11 (4 professional)
Architects 0 2 1 1
Artists (other) 2 5 4 5
Information determined via Performance Background Questionnaire. Statistics and frequencies generated via SPSS software.
57
Table 2: Population Characteristics, Architecture Students with Significant Musical Training (AWM)
Freshmen (n=3) Seniors (n=5)
Number of women 1 2
Number of men 2 3
Age at scan time 18.67 (± 1.15) 22.4 (± 0.55)
Subjects with absolute pitch 1 0
Age of onset, musical training
First instrument 6.00 (± 3.61) 9.20 (± 7.19)
Primary instrument 10.00 (± 5.29) 15.00 (± 6.16)
Secondary instrument 8.67 (± 5.13) 8.00 (± 4.69)
Years played, musical instrument
Primary instrument 8.50 (± 5.22) 5.80 (± 3.70)
Secondary instrument 8.67 (± 6.03) 9.00 (± 6.06)
Intensity of practice (avg. hrs./day), musical instrument
Primary instrument 1.67 (± 1.15) 0.52 (± 0.37)
Secondary instrument 1.12 (± 1.24) 0.63 (± 0.32)
Subjects with pre-college musical training
Music theory 0 1
Music history 0 0
Aural skills (ear training) 0 1
Subjects with pre-college architectural training 1 3
Subjects with relatives in the arts 3 4
Musicians 2 (1 professional) 3 (1 professional)
Architects 1 3
Artists (other) 2 3
Information determined via Performance Background Questionnaire. Statistics and frequencies generated via SPSS software.
58
Table 3: Population Differences, by Major
ARC vs.
Mean
Diff.
MUS*
Std. Err.
p-value
ARC vs.
Mean
Diff.
AWM**
Std. Err.
p-value
MUS vs.
Mean
Diff.
AWM***
Std. Err.
p-value
Age of onset, musical training
First instrument 2.95833 0.58077 0.000** 0.25000 2.18162 0.912 -2.70833 2.13958 0.245
Primary instrument 0.63768 1.08264 0.561 -3.79167 2.28208 0.131 -4.42935 2.25098 0.082ᵗ
Secondary instrument 3.36364 1.23932 0.012* 4.21429 1.86719 0.052ᵗ 0.85065 1.93303 0.669
Years played, musical instrument
Primary instrument -7.84058 0.97112 0.000** -3.47917 1.55340 0.054ᵗ 4.36141 1.70932 0.025*
Secondary instrument -5.69048 1.18043 0.000** -5.85714 2.12892 0.031* -0.16667 2.35979 0.945
Intensity of practice (average hours/day), musical
instrument
Primary instrument -1.03174 0.38032 0.021* -0.11429 0.45978 0.809 0.91745 0.43132 0.054ᵗ
Secondary instrument 0.34733 0.47124 0.522 0.50333 0.54392 0.406 0.15600 0.35919 0.678
Information determined via Performance Background Questionnaire. Statistics and frequencies generated via SPSS software. Equal variances not assumed.
* Architecture (1) vs. Music (2); a negative mean value indicates a higher mean in group 2 (Music).
** Architecture (1) vs. AWM (2); a negative mean value indicates a higher mean in group 2 (AWM).
*** Music (1) vs. AWM (2); a negative mean value indicates a higher mean in group 2 (AWM).
59
Table 4: Population Differences, by Major and Year
ARCF vs.
Mean
MUSF*
Std. Err.
p-value
ARCS vs.
Mean
MUSS**
Std. Err.
p-value
Age of onset, musical training
First instrument 3.80000 1.11555 0.022* 2.35714 0.64000 0.002**
Primary instrument 1.80000 1.65193 0.303 -0.23077 1.49240 0.879
Secondary instrument 4.13889 1.96987 0.063ᵗ 2.90385 1.66688 0.109
Years played, musical instrument
Primary instrument -5.50000 1.60797 0.005** -9.61538 1.08500 0.000**
Secondary instrument -6.11111 1.56742 0.003** -5.37500 1.76361 0.009**
Intensity of practice (average
hours/day), musical instrument
Primary instrument -1.19333 0.44158 0.021* -0.73692 0.81173 0.495
Secondary instrument 0.56333 0.31503 0.134 0.23111 0.79801 0.812
Information determined via Performance Background Questionnaire. Statistics and frequencies generated via SPSS software.
Equal variances not assumed.
* Architecture Freshmen (1) vs. Music Freshmen (2); a negative mean value indicates a higher mean in group 2 (Music Freshmen).
** Architecture Seniors (1) vs. Music Seniors (2); a negative mean value indicates a higher mean in group 2 (Music Seniors).
60
Table 5 reveals the instrumentation breakdown by major and by year. As can be
expected, all music majors are currently playing their primary instruments, and 9/10
freshmen and 9/15 seniors, respectively, are playing multiple instruments. That more
freshmen than seniors are playing multiple instruments can either be attributed to the
increased time demand for the primary instrument placed on seniors, who are typically
performing professionally, or may simply be attributed to chance in this population. As
can be seen on the table, 13/15 a majority of music seniors have had a secondary
instrument in the past. As mentioned above, no architecture majors are currently playing
any instrument, although 5/6 freshmen and 7/10 seniors had some form of musical
engagement in the past, however brief.
2. Population Characteristics by Instrument Group
Tables 6-7 detail the same population characteristics and statistical comparisons
as those found in Tables 1-4, but for three subgroups of the music major population.
Music majors were recruited as three target groups: piano majors, strings majors, and
voice majors. For the purposes of this analysis, they were grouped by instrument, and not
by years in school. Table 6 reveals that: the piano group (n=6), has 3 women and 3 men,
and 3 seniors and 3 freshmen; the strings group (n=9), has 5 women and 4 men, and 4
seniors and 5 freshmen; the voice group (n=10), has 7 women and 3 men, and 8 seniors
and 2 freshmen.
The difference in pre-collegiate academic musical training between the three
groups is apparent: while 5/6 piano majors and 8/9 strings majors underwent music
theory training prior to entering conservatory, only 5/10 voice majors did so.
Interestingly, 7/10 voice majors underwent ear training (usually solfege), which is similar
61
to the number of piano majors (3/6) and strings majors (4/9) that underwent some form of
academic ear training prior to entering conservatory. Aural skills (ear training) are
particularly crucial for vocalists, who cannot rely on instrument fingerings or bow
positions to produce accurate pitches.
In Table 7, we see that piano majors had a statistically significant younger age of
onset of musical training of their primary instrument than either the string players or
voice majors. There is no statistically significant difference among the three groups for
age of onset of musical training for the first instrument, however; the piano majors
typically started on and stayed with piano, while the string and voice majors often started
with piano (or another stringed instrument), and later added what became their primary
instrument. This same holds for the difference in years played (primary instrument) for
piano majors compared to both strings and voice majors; there is a very statistically
significant difference between piano majors and strings majors, and between piano
majors and voice majors for this variable. The most interesting results presented on this
table correspond to the intensity of practice, for both piano and string majors compared to
voice majors. Piano majors practice significantly longer each day than voice majors
(p=0.011), and strings majors show a strong statistical trend toward longer daily practice
than voice majors (p=0.059). While this last difference can be explained by the fact that
the use of the vocal cords simply cannot be tolerated for the same amount of time per day
without damage, one wonders how much the differing levels of intensity of practice
among these three groups may have modulated the expected neuroplastic changes in
these musician’s brains, particularly in motor-related areas like the precentral and
postcentral gyri, and the cerebellum.
62
Table 5: Population Characteristics, Instrumentation
Architecture
Freshmen (n=6)
Seniors (n=10)
Music
Freshmen (n=10)
Seniors (n=15)
AWM
Freshmen (n=3)
Seniors (n=5)
Currently playing primary instrument 0 0 10 15 3 4
Currently playing secondary instrument 1 0 8 9 2 1
Currently playing multiple instruments 0 0 9 9 2 2
Primary Instrument 5 7 10 15 3 5
Piano 3 6 3 3 1 2
Violin 1 0 1 2 0 0
Viola 0 0 2 1 0 0
Cello 0 0 1 1 0 0
Double Bass 0 0 1 0 0 0
Guitar 0 0 0 0 1 2
Flute 0 1 0 0 0 1
Trombone 1 0 0 0 0 0
Voice 0 0 2 8 0 0
Other 0 0 0 0 1 0
Secondary instrument 4 4 9 13 3 4
Piano 0 0 5 5 0 1
Violin 1 0 1 4 0 1
Viola 0 0 0 1 0 0
Cello 0 0 1 0 0 0
Guitar 0 1 0 0 0 0
Clarinet 1 1 0 0 0 0
Oboe/English Horn 0 0 0 2 0 0
Percussion 0 0 0 1 1 0
Voice 2 0 2 0 2 2
Other 0 2 0 0 0 0
Information determined via Performance Background Questionnaire. Statistics and frequencies generated via SPSS software. Only relevant instrument
classifications are listed for the primary and secondary instrument categories; for example, as no subject listed oboe or English horn as their primary instrument,
the classification was removed from the table.
63
Table 6: Population Characteristics, by Instrument Group (Music Majors)
Piano (n=6) Strings (n=9) Voice (n=10)
Number of women 3 5 7
Number of freshmen 3 5 2
Age at scan time 20.00 (± 2.19) 19.67 (± 1.58) 21.10 (± 1.92)
Subjects with absolute pitch 0 2 1
Age of onset, musical training
First instrument 5.00 (± 0.63) 4.87 (± 1.13) 5.80 (± 1.75)
Primary instrument 5.17 (± 0.75) 9.87 (± 2.47) 10.00 (± 3.77)
Secondary instrument 10.00 (± 3.92) 9.00 (± 4.11) 8.90 (± 5.22)
Years played, musical instrument
Primary instrument 14.83 (± 2.14) 8.50 (± 4.75) 11.11 (± 2.47)
Secondary instrument 6.50 (± 4.51) 7.62 (± 3.96) 10.61 (± 5.89)
Intensity of practice (average hours/day), musical instrument
Primary instrument 2.53 (± 0.81) 2.32 (± 1.29) 1.18 (± 0.92)
Secondary instrument 0.75 (± 0.00) 0.90 (± 0.37) 1.07 (± 0.80)
Subjects with pre-college musical training
Music theory 5 8 5
Music history 1 1 4
Aural skills (ear training) 3 4 7
Subjects with pre-college architectural training 1 0 0
Subjects with relatives in the arts 4 6 9
Musicians 4 (1 professional) 6 (3 professional) 9 (3 professional)
Architects 0 1 1
Artists (other) 2 4 3
Information determined via Performance Background Questionnaire. Statistics and frequencies generated via SPSS software.
64
Table 7: Population Differences, by Instrument Group (Music Majors)
Piano vs.
Mean
Diff.
Strings*
Std. Err.
p-value
Piano vs.
Mean
Diff.
Voice**
Std. Err.
p-value
Strings
vs.
Mean
Diff.
Voice***
Std. Err.
p-value
Age of onset, musical training
First instrument 0.12500 0.51412 0.797 -0.80000 0.61101 0.214 -0.92500 0.68202 0.195
Primary instrument -4.70833 0.92740 0.001** -4.83333 1.29529 0.005** -0.12500 1.53263 0.936
Secondary instrument 1.00000 2.43731 0.695 1.10000 2.56016 0.680 0.10000 2.19733 0.964
Years played, musical instrument
Primary instrument 6.33333 1.89276 0.007** 3.72222 1.20005 0.009** -2.61111 1.87095 0.192
Secondary instrument -1.12500 2.65431 0.688 -4.11111 2.99047 0.208 -2.98611 2.41280 0.236
Intensity of practice (average hours/day), musical
instrument
Primary instrument 0.20667 0.56356 0.720 1.34944 0.45017 0.011* 1.14278 0.55037 0.059ᵗ
Secondary instrument -0.15500 0.18558 0.465 -0.32444 0.26819 0.261 -0.16944 0.32614 0.614
Information determined via Performance Background Questionnaire. Statistics and frequencies generated via SPSS software. Equal variances not assumed.
* Piano (1) vs. Strings (2); a negative mean value indicates a higher mean in group 2 (Strings).
** Piano (1) vs. Voice (2); a negative mean value indicates a higher mean in group 2 (Voice).
*** Strings (1) vs. Voice (2); a negative mean value indicates a higher mean in group 2 (Voice).
65
The fact that I created a middle group (AWM), together with the existence of
statistically significant differences in primary instrument age of onset of musical training,
years played, and intensity of practice between instrument groups within the music major
population, merits the exploratory investigation of the dataset within the music majors
based on instrument group, and between the middle group (AWM) and both architecture
majors and music majors, in addition to the planned investigation of the dataset based on
both type of major (architecture or music) and years of study (freshman or senior). The
following chapters detail these planned and exploratory analyses within different
methodological approaches of the neuroanatomical data.
66
Performance Background Questionnaire
1. Please indicate the following for your own musical history: (a) what musical
instruments have you learned to play at any point, (b) what musical instruments do you
currently play, (c) how long has each instrument been played, (d) how old were you when
you started learning each instrument, (e) what instruments have you taken private lessons
on, and (f) what instruments have you participated in ensembles with. Please also indicate
what you consider to be your primary (P) and secondary (S) instruments.
Instrument Ever
played?
Yes/No
Currently
play?
Yes/No
# of
years
played
Age
instrument
learning
started
Private
lessons?
Yes/No
Ensemble
participation?
(Please
indicate type
of ensemble)
Piano
Violin
Viola
Cello
Bass (upright)
Flute
Clarinet
Oboe/Eng. Horn
Bassoon
Recorder
Trumpet
French Horn
Trombone
Tuba
Percussion
Guitar
Voice
Other (Specify):
_____________
67
1b. If you indicated above that you participated in an ensemble, please also indicate
above the number of years of participation for each type of ensemble.
1c. For your primary and secondary instruments, how many hours a day on average have
you practiced each year since you began learning the instrument? If there is a year-by-
year difference, please indicate.
2. Have you ever participated in marching band? If so, for how many years, and what
instrument do/did you play for it?
3. Are any of your relatives musicians? If so, are they professional musicians, and what
instruments do they play?
4. What type (if any) of music theory, history, and/or aural skills did you study before
college (ex: Kodaly method, solfège, AP music theory, etc.)?
5. Do you have perfect pitch?
6. Do you have any pre-college architectural experience (summer programs, etc.)?
7. Are any of your relatives architects?
8. Are any of your relatives artists of any type (actors, film-makers, painters, sculptors,
architects, musicians, etc.)?
68
CHAPTER 3
A. BACKGROUND
As discussed in the first chapter, it remains unclear as to why there are so many
differences in neuroanatomical results between musicians and non-musicians across
studies, in spite of the similarities in populations. These results are presented in Table 1,
parsed by instrumentation and gender.
This summary table makes it clear that the most consistent areas of grey matter
density differences for musicians when compared to non-musicians are sensory and
motor areas. Regions of interest repeatedly identified (across at least two of the discussed
studies) are primary motor cortex (precentral gyrus), somatosensory cortex (postcentral
gyrus), Heschl’s gyrus, inferior frontal gyrus (largely on the left), and cerebellum. As the
populations studied almost always included male keyboardists and strings players, but
less often females or other types of instruments, studies that use gender and
instrumentation as covariables are needed to help identify neuroanatomical differences
generalizable to all musicians. While the longitudinal study mentioned in chapter one has
much to offer, it is limited by the fact that the only instruments the subjects are playing
are keyboard and strings instruments. It will be interesting to see how their brains change
when, in a decade’s time, some choose to switch instruments or stop playing altogether.
This longitudinal investigation can be greatly enhanced by cross-sectional and
longitudinal studies in adults (particularly adults undergoing conservatory training or the
learning of new instruments), and in both trained and novice musicians.
In this chapter I report the results of the investigation of grey matter differences
detected using voxel-based morphometry (VBM), the method preferred in the studies in
69
Region of Interest Keyboardists Mixed Instrumentation
Precentral Gyrus
3
●
Postcentral Gyrus
1
● ◐
“Premotor Areas”
1
●
Heschl’s Gyrus
1
◐ ◐
● ●
Inferior Frontal Gyrus
1
◐ ◐ ◐
Cerebellum
1
● ●
Parietal Lobe
12
● ◐
Prefrontal Cortex
●
Planum Temporale
◑ ◑
Planum Polare
◑ ◑
Hippocampus
◐ ◐ ◐ ◐
Fusiform Gyrus
◑ ◑
Middle Orbitofrontal Gyrus
◑ ◑
Table 1. Areas of increased gray matter density found via voxel-based morphometry
analyses of musicians as compared to non-musicians.
● = bilateral
◐ = left hemisphere
◑ = right hemisphere
1
Same results (as column 1) found for amateur male keyboardists
2
Left hemisphere results are in the intraparietal sulcus
70
the extant literature, summarized in Table 1. VBM is a widely utilized automated
structural analysis technique that yields results easily and within a modest amount of
time. VBM operates without need for a priori defined regions of interest (ROIs), and thus
can be used as a tool that will yield voxelwise results over the entire brain. For the
purposes of this study, VBM was used at the level of the whole brain, and for increased
statistical power, also within a priori defined regions of interest.
B. METHODS
i. Subjects
Fifty two subjects participated in the study; 2 were excluded because of incidental
findings, and 1 was excluded because the age at scan time exceeded my pre-established
age window. Eight were further removed (architecture group) due to previous significant
musical training. The remaining 41 subjects formed four primary groups: music major
freshmen (n=10, 5 females), music major seniors (n=15, 10 females), architecture major
freshmen (n=6, 3 females), and architecture major seniors (n=10, 5 females). Musicians
represented three instrument families: strings, keyboard, and voice. Performance
background information was collected via questionnaire (please see chapter 2, page 66);
for more details on the subject population, please see chapter 2.
ii. Image Acquisition
High-resolution anatomical MRI and Diffusion MRI scans were obtained in all
subjects using a Siemens 3-Tesla scanner. For the structural MRI data, a high resolution
T1-weighted MPRAGE sequence with the following parameters was used: TR = 2530ms,
71
TE = 3.09ms, FOV = 256mm x 256mm, 208 slices, Flip angle = 10, Matrix = 256 x 256,
Slice thickness = 1mm. A 64-direction diffusion weighted imaging sequence was
performed next using the following parameters: TR = 10000ms, TE = 100ms, 70 slices,
slice thickness = 2.5mm, FOV = 224mm, Matrix = 128 x 128. Voxel-based morphometry
analyses were performed on all MPRAGE data using FSL software.
iii. Analysis
Automated VBM analysis was performed (Ashburner, 2000, Good, 2001) using
FSL (Smith, 2004). Structural images were renamed, converted to NIfTI format using
DCM2NII software, averaged (both MPRAGE scans) to reduce the signal-to-noise ratio,
and finally extracted using BET (Smith, 2002). Many of the basic/automatic brain
extractions still included meninges, dura, skull, neck, and other extra-brain tissues. To
solve this issue, BET was repeated with different thresholds until the best possible
automated extraction was achieved for each brain. Tissue-type segmentation was
performed using FAST4 (Zhang, 2001). Resulting grey matter partial volume images
were aligned to MNI152 standard space using FLIRT, an affine registration tool
(Jenkinson, 2001, 2002). Nonlinear registration was performed using FNIRT (Andersson,
2007a, 2007b), which uses a b-spline representation of the registration warp field
(Rueckert, 1999). The resulting images were averaged to create a study-specific template,
to which the native grey matter images were non-linearly re-registered. The registered
partial volume images were modulated (to correct for local expansion or contraction) by
dividing by the Jacobian of the warp field. The modulated segmented images were
smoothed with an isotropic Gaussian kernel, and finally, voxelwise GLM was applied
72
using permutation-based non-parametric testing, to correct for multiple comparisons
across space. In this way, data from multiple subjects was comparable on a voxelwise
basis. These voxelwise findings were used to address ROI predictions (see chapter 2,
pages 5-6). Due to the nature of VBM methods (unbiased by a priori hypotheses), the
process sometimes yields unpredicted differences between subject groups. Although I
expected to find reliable results in the predicted ROIs, this was not the case.
C. PROPOSED/ORIGINAL ANALYSES
i. Results
VBM analyses were performed in two stages, summarized in Table 2. Stage 1
made voxelwise comparisons across the whole brain between four groups: music seniors,
music freshmen, architecture seniors, and architecture freshmen. Stage 2 made voxelwise
comparisons within each of the a priori defined ROIs, to improve statistical power.
Statistically significant and statistical trend results are reported within each Stage, by
group comparison. All peak voxel coordinates are reported in X,Y,Z format.
a. Stage 1
Stage 1 compared four groups: music seniors, music freshmen, architecture
seniors, and architecture freshmen. Statistically significant voxelwise differences were
found: in music seniors as compared to music freshmen in right cerebellum (31, 27, 22;
p=0.0474*); and in music freshmen as compared to music seniors in L angular
gyrus/inferior parietal lobule (66, 35, 58; p=0.0422*).
73
Table 2: Voxel-Based Morphometry Results, Planned Analysis (Major and Year)
Comparison Hemisphere Result p-value
Music Majors > Architecture Majors Left Anterior Insula p=0.0362*
Architecture Majors > Music Majors N/A
Music Seniors > Architecture Seniors N/A
Architecture Seniors > Music Seniors Right Precuneus/
Posterior Cingulate
Gyrus
p=0.0212*
Music Freshmen > Architecture Freshmen N/A
Architecture Freshmen > Music Freshmen N/A
Music Seniors > Music Freshmen Right Cerebellum p=0.014*
Music Freshmen > Music Seniors Left Angular Gyrus p=0.0422*
Architecture Seniors > Architecture Freshmen N/A
Architecture Freshmen > Architecture Seniors Left PTR/Orbitofrontal
Cortex
p=0.0234*
Left POP/Precentral
Gyrus
p=0.0428*
Left Anterior Cingulate
Gyrus
p=0.0824ᵗ
Left Superior/Posterior
Cerebellum
p=0.0818ᵗ
All data analyzed and statistics processed via FSL. Table shows only very statistically significant (**),
statistically significant (*), or statistical trend (ᵗ) results.
Statistical trends toward significance were found: in architecture seniors as compared to
music seniors in right precuneus/superior parietal lobule (42, 34, 58; p=0.076).
b. Stage 2
Stage 2 compared the same four groups. Here the analysis was restricted to each
of the following ten a priori defined ROIs in turn: cerebellum, cingulate gyrus, frontal
lobe, Heschl’s gyrus, hippocampus, inferior frontal gyrus (both the pars opercularis
(POP) and the pars triangularis (PTR)), insula, postcentral gyrus, and precentral gyrus.
Because the amount of tissue to be interrogated is smaller than the initial whole brain
74
analysis, performing it in this manner increases the statistical power of detection in each
of the ROIs.
Statistically significant voxelwise differences were found: in architecture seniors as
compared to music seniors in R posterior cingulate gyrus (42, 35, 57; p=0.0212*); in
architecture freshmen as compared to architecture seniors in L precentral gyrus (67, 63,
49; p=0.0428*), and L orbitofrontal gyrus (17, 79, 30; p=0.0234*); in music seniors as
compared to music freshmen in R cerebellum (30, 27, 23; p=0.014*); and in all music
majors as compared to all architecture majors in L anterior insula (61, 72, 37;
p=0.0362*).
Statistical trends toward significance were found: in architecture freshmen as compared
to architecture seniors in L cerebellum (62, 26, 22; p=0.0818), and L anterior cingulate
gyrus (48, 84, 40; p=0.0824).
ii. Discussion
Despite the statistical significance – and trends toward significance - of these
results, one cannot forget the fact that the groups are small, between 6-15 subjects each.
Given the small numbers, strong differences in individual subjects (in reality outliers)
could be driving some of the findings; I will discuss this possibility when I address group
comparisons.
Beginning with the comparison between all architecture majors and all music
majors, only one voxelwise difference surfaced. Although no results were found in most
75
of the hypothesized regions of interest, the finding of increased grey matter in left
anterior insula for music majors compared to architecture majors does fit my predictions.
As reviewed by Craig (2009), the anterior insular cortex (AIC) has been linked in
functional imaging studies to emotion, at the individual and group levels, self-awareness,
object identification, attention, and different aspects of musical processing: rhythmic
processing (L AIC; Platel, 1997) and listening to subjectively pleasant music (L AIC;
Koelsch et al, 2006). Additionally, the AIC has recently been functionally implicated in
feedback necessary for vocal control in singers (Zarate, 2013), and increased empathy has
been positively correlated with grey matter density in the left AIC (Mutschler et al,
2013). The population of music majors within the present study is comprised of 10 voice
majors, 9 strings majors, and 6 piano majors. While functional usage of the left AIC for
rhythmic processing and listening to subjectively pleasant music is valid for the entire
music major population, it stands to reason that the relatively high number of vocalists
might be biasing the results, given the functional involvement of the AIC in vocal control
in singers. That said, group level emotional awareness and empathy are important for
both instrumental and vocal group music-making, such as orchestral and choral ensemble
participation. Students in conservatory are typically required to participate in ensembles
of differing sizes throughout their training, theoretically honing an emotional and musical
skill set that likely demands increased functional involvement of the AIC. Four years of
heavy engagement (increased over pre-conservatory involvement) would likely be
reflected structurally, thus supporting these results.
When the comparison of music majors and architecture majors is split by years of
study, the only statistically significant difference found is that of increased grey matter in
76
senior architecture majors as compared to senior music majors in right posterior cingulate
gyrus. This is compounded by a statistical trend found in neighboring right precuneus.
The posterior cingulate gyrus has been linked in fMRI studies to aesthetic judgment
(Jacobsen et al, 2006, 2010). One anatomical study found that an increase in cortical
thickness in the posterior cingulate positively correlated with scores on creative thinking
tests. The precuneus has been implicated in both imagery and memory retrieval tasks
(Fairhall & Ishai, 2008). Taken together, these references lead one to surmise that the
increased creative thought and output of architecture majors might have led, over the
course of four years of conservatory-style training, to the small, yet statistically
significant voxelwise differences found between architecture seniors and music seniors in
the right posterior cingulate/precuneus region. This is not to say that music majors are not
creative, but simply that the design-based training these architecture students are
undergoing is perhaps more functionally demanding – and hence, structurally changing –
than the training undergone by these music majors.
Statistically significant differences – and trends toward significance – were found
when comparing years of study (freshmen or seniors) within major disciplines. First,
fitting with the hypotheses of this study, senior music majors were found to have
increased grey matter in right cerebellum when compared to freshmen music majors. This
prediction was made based on the cerebellum’s known role in motor movement, and the
finding of larger cerebellar volume and grey matter in pianists as compared to non-
musicians (Gaser & Schlaug, 2003, Hutchinson et al, 2003, Han et al, 2009, Giedd &
Rapoport, 2010). My results, lateralized to the right hemisphere, corroborate those of Han
and Hutchinson, but not those of Gaser & Schlaug, who only found left hemisphere
77
increases in cerebellum. Additionally, the cerebellum has been implicated in aesthetic
judgment of music, particularly rhythmic music (Brattico et al, 2013). Aesthetic
judgment of one’s own artform is a skill that begins to develop early in life, but is
deliberately honed during conservatory training; thus, it makes sense that this honing
might increase grey matter in the right cerebellum of senior music students as compared
to freshmen.
There were no statistically significant results or trends toward significance for
architecture seniors as compared to architecture freshmen. This seems odd, given the four
years of intensive training. There were, on the other hand, significant grey matter
increases found for architecture freshmen when compared to architecture seniors, as well
as for music freshmen when compared to music seniors. For the architecture major
comparison, these differences were found entirely in the left hemisphere: results were
statistically significant in precentral gyrus and in orbitofrontal gyrus, and statistical trends
were found in anterior cingulate gyrus, and cerebellum. The precentral gyrus results
(Figure 1) are somewhat inferior and deep within the sulcus; the representative motor
function of this area is difficult to determine, given that the results appear to lie in white
matter. The orbitofrontal results are quite anterior, indeed bordering on the frontal pole.
The orbitofrontal cortex has been functionally linked to reward learning, decision
making, aesthetic appreciation, and aesthetic judgment (Nadal 2008, Kirk 2009, Jacobsen
2010, Chakravarty 2012, Kahnt 2012), all important functions both in architects and in
musicians as they hone their crafts during the course of their conservatory-style training.
78
Figure 1: Architecture freshmen as compared to architecture seniors in precentral gyrus
(p=0.428*).
Thus, the finding of increased grey matter in orbitofrontal cortex for architecture
freshmen when compared to architecture seniors, and not the other way around, seems
odd. While these results are interesting, it cannot be overlooked that much of the frontal
lobe is late to fully mature, and that grey matter density declines in this region through
young adulthood (Gogtay et al, 2004, Thompson et al, 2005, Whitford et al, 2007); it is
thus possible that this greater grey matter density in architecture freshmen when
compared to seniors is driven by a developmental change that has occurred in the latter
population, but not yet in the former population. This line of reasoning would also
explain larger grey matter density in freshmen as compared to seniors in both
orbitofrontal cortex and anterior cingulate gyrus. If truly driven by development,
however, one would also expect to see these results in music freshmen when compared to
music seniors; the lack of these differences in that group comparison suggests that either
79
the architecture freshmen results are owed to something else, or that the music freshmen
are either somehow further along in development, or have had experience-dependent
modulation of these frontal brain regions. Increased grey matter in left cerebellum in this
same population is less easy to explain. As discussed above, the cerebellum was
predicted to have increased grey matter in music majors as compared to architecture
majors because of its known role in motor movement; while many architecture majors in
this population have experienced some form of musical training prior to college, there is
not a significantly higher number of subjects with rhythmic instrumental experience in
either the freshman or senior groups (see chapter 2, table 5 for instrument distribution).
Background information regarding other motor-related experiences, such as dance or
athletic training that each subject might have undergone was, unfortunately, not collected.
That these differences are not present in the seniors either indicates that the students are
beginning to conform to adult gender differences (i.e. the findings are related to a purely
developmental issue), or that, given that this study is cross-sectional, there is a truly
qualitative difference between these architecture freshmen and seniors, attributable to
other individual differences in environment.
Grey matter increases were found in freshmen music majors as compared to
senior music majors in left angular gyrus. The left angular gyrus has been functionally
linked to both language and mathematical skill (Dehaene et al, 1999, Grabner et al,
2007). Potentially more relevant to this particular population, it has also been suggested
that it serves a role in metaphor comprehension and certain forms of synaesthesia
(Hubbard & Ramachandran, 2003). Although we have no information regarding
synaesthesia in this population, given its higher incidence in artistic populations (Brang
80
& Ramachandran, 2011), it is possible that the presence of more synaesthetes could be
driving this difference. It is more likely, however, that, as suggested in the previous
paragraph, there is simply a qualitative difference in these 10 freshmen music majors as
compared to the 15 senior music majors. This could be due to mathematical training in
architects (music majors have no math requirements during conservatory), or to
pronounced individual differences (outliers) driving the results.
Given that, as demonstrated in chapter 2, there are several potential subgroups
within the architecture and music populations, and that there is a middle group,
architecture majors with significant musical training (AWM), it was deemed pertinent to
conduct several exploratory analyses of the dataset.
D. EXPLORATORY ANALYSES
1. Instrumentation
i. Results
VBM analyses were performed in two stages, as before, summarized in Table 3. Stage 1
made voxelwise comparisons across the whole brain between five groups: piano majors,
strings majors, voice majors, instrumentalists (both piano and strings majors; compared
to voice majors), and architecture seniors, chosen to represent the architecture population.
Stage 2 made voxelwise comparisons within each of the a priori defined ROIs, to
improve statistical power. Statistically significant and statistical trend results are reported
within each Stage, by group comparison. All peak voxel coordinates are reported in
X,Y,Z format.
81
Table 3: Voxel-Based Morphometry Results, Exploratory Analysis 1 (Instrumentation)
Comparison Hemisphere Result p-value
Instrumentalists > Vocalists Left Planum Temporale/
Superior Temporal
Gyrus
p=0.0182*
Left PTR/Frontal Pole p=0.0208*
Left Insula/Heschl’s Gyrus p=0.0734ᵗ
Vocalists > Instrumentalists Right POP p=0.0954ᵗ
Piano Majors > Strings Majors N/A
Strings Majors > Piano Majors Right Planum Polare/
Heschl’s Gyrus
p=0.0268*
Piano Majors > Voice Majors Left Insula/Heschl’s Gyrus p=0.0596ᵗ
Voice Majors > Piano Majors N/A
Strings Majors > Voice Majors Left Planum Temporale/
Superior Temporal
Gyrus
p=0.0356*
Left PTR p=0.0068**
Right Anterior Cingulate
Gyrus/ Paracingulate
p=0.0734ᵗ
Voice Majors > Strings Majors Left Anterior Insula p=0.0570ᵗ
Right Lateral Precentral Gyrus p=0.0888ᵗ
Instrumentalists > Architecture Majors N/A
Architecture Majors > Instrumentalists N/A
Vocalists > Architecture Majors Left Anterior Insula p=0.0422*
Architecture Majors > Vocalists Left PTR/Frontal Pole p=0.0976ᵗ
Piano Majors > Architecture Majors N/A
Architecture Majors > Piano Majors Left Precuneus/
Posterior Cingulate
Gyrus
p=0.0842ᵗ
Strings Majors > Architecture Majors N/A
Architecture Majors > Strings Majors N/A
All data analyzed and statistics processed via FSL. Table shows only very statistically significant (**),
statistically significant (*), or statistical trend (ᵗ) results. Instrumentalists group is comprised of both
freshman and senior piano majors and strings majors; Vocalists group is comprised of both freshman and
senior voice majors. Architecture Majors group is comprised of senior architecture majors, to keep similar
group sizes. Piano, Strings, and Voice major groups include both freshmen and seniors.
a. Stage 1
Stage 1 compared five groups: piano majors, strings majors, voice majors,
instrumentalists (both piano and strings majors; compared to voice majors), and
architecture seniors. Subjects were not split by year. No statistically significant voxelwise
differences or statistical trends were found in this analysis.
82
b. Stage 2
Stage 2 compared the same five groups, within each of the ten a priori defined
ROIs. Statistically significant voxelwise differences were found: in strings majors when
compared to piano majors in R planum polare/HG (20, 59, 33; p=0.0268*); in strings
majors when compared to voice majors in L planum temporale/superior temporal gyrus
(74, 53, 39; p=0.0356*), and in L PTR (71, 80, 37; p=0.0068**); in voice majors when
compared to architecture seniors in L anterior insula (60, 72, 35; p=0.0422*); in
instrumentalists as compared to voice majors in L planum temporale/superior temporal
gyrus (70, 52, 35; p=0.0182*), and in L PTR/frontal pole (70, 82, 42; p=0.0208*).
Statistical trends toward significance were found: in architecture seniors when compared
to piano majors in L precuneus/posterior cingulate gyrus (47, 34, 50; p=0.0842); in
architecture seniors when compared to voice majors in L PTR/frontal pole (73, 80, 35;
p=0.0976); in piano majors when compared to voice majors in L posterior insula/HG (61,
51, 39; p=0.0596); in strings majors when compared to voice majors in R paracingulate
gyrus/anterior cingulate gyrus (41, 78, 52; p=0.0734), in L PTR (70, 79, 40; p=0.0634);
in voice majors when compared to strings majors in L anterior insula (61, 67, 32;
p=0.057), and in R lateral precentral gyrus (15, 65, 52; p=0.0888); in instrumentalists
when compared to voice majors in L posterior insula/HG (61, 51, 40; p=0.0734); and in
voice majors when compared to instrumentalists in R POP (20, 69, 42; p=0.0954).
83
ii. Discussion
Results from this exploratory analysis reveal differences in key regions of interest
between instrumentalists (piano and strings, both separately and together) and vocalists.
One would expect these differences to fall in precentral and postcentral gyri, particularly
in the hand knob and related areas, given the different motor demands of singing, playing
the piano, or playing a stringed instrument. However, the only reportable result in either
of these regions was a statistical trend found in right lateral precentral gyrus in voice
majors as compared to strings majors. Perhaps the general lack of difference in these
regions is due to the fact that most voice majors have played or currently play piano; it is
often a skill necessary for teaching private voice lessons, something that many voice
majors plan to do after graduation from conservatory.
More surprising are the findings of differences in both auditory cortex and inferior
frontal gyrus. Specifically, all instrumentalists had increased grey matter in three left
hemisphere regions when compared to vocalists: planum temporale/superior temporal
gyrus, PTR/frontal pole, and insula/Heschl’s Gyrus. The first two results were
statistically significant, while the latter was a statistical trend. The planum
temporale/STG result was driven by the strings majors, and the insula/HG result was
driven by the piano majors (and was more significant in the piano > voice comparison).
When compared to the instrumentalists, the vocalists exhibited a statistical trend toward
increased grey matter in right POP. Without the ability to investigate individual
differences within each group, these results are difficult to explain convincingly. It is
possible that the increased intensity of practice and/or the increased amount of pre-
84
college music theory training in both piano majors and strings majors compared to voice
majors (see chapter 2, table 6) might have had an effect on planum temporale, HG, and/or
parts of inferior frontal gyrus, as that increased training would certainly have included ear
training (through instrumental practice) and the study of musical scores, involving
processing of musical syntax and semantics. However, results in planum temporale and
HG due to pre-conservatory levels of ear training should be present in these voice majors
as well; many of them underwent ear training (solfege) prior to entering conservatory.
As there are differences present among instrument groups, it may be expected that
there will also be some differences between those instrument groups – specifically the
voice majors – and architecture majors. In fact, there is statistically significant increased
grey matter in left anterior insula in voice majors when compared to architecture majors.
This same result, in the form of a strong statistical trend, also is there in voice majors
when compared to strings majors. As discussed in the planned analysis section, before,
the left anterior insula has been implicated as contributing to vocal control in singers
(Zarate, 2013). I suggested that the voice majors might be biasing the results between
music majors and architecture majors, and indeed, this last result makes that explanation
more likely. The functional involvement of the left anterior insula in vocal control would
also likely result in voxelwise differences between vocalists and both strings and piano
majors, although there were no differences between voice majors and piano majors to
report.
Interestingly, there are two statistical trends toward increased grey matter in
architecture majors over instrument groups: in architecture majors when compared to
piano majors in left precuneus/posterior cingulate gyrus, and in architecture majors when
85
compared to voice majors in left PTR/frontal pole. The latter is, oddly, roughly the same
result as found in all instrumentalists when compared to all vocalists. This suggests that
there is a section of left PTR/frontal pole within this voice major population that is
simply less dense in terms of grey matter than the same region in the rest of this
population, likely attributable to individual differences. The left precuneus/posterior
cingulate gyrus result in architecture majors when compared to piano majors is in the
opposite hemisphere to the (statistically significant) architecture seniors as compared to
music seniors result, seen in the planned analysis, above. In addition to the posterior
cingulate gyrus and precuneus functions already discussed, Kowatari (et al, 2009) found
a functional increase in the left parietal lobe in artists trained in design versus novices
when performing a creative design task. Since the architecture program these students are
undergoing is heavily based on design, it is not surprising that creative thought and
design related brain regions might exhibit increased grey matter when compared to the
same regions in music majors. That this result is only present in the comparison to piano
majors is, however, intriguing and perhaps more difficult to explain.
The differences among instrument groups could help explain the general lack of
differences found between music majors and architecture majors in the planned analysis
discussed above.
The next exploratory analysis conducted examined possible gender effects in the
dataset.
86
2. Gender
i. Results
VBM analyses were performed, once more, in two stages, summarized in Table 4.
In Stage 1, voxelwise comparisons were made across the whole brain between six groups:
architecture females, architecture males, music females, music males, and all females and
all males. Stage 2 made voxelwise comparisons within each of the a priori defined ROIs,
to improve statistical power. Statistically significant and statistical trend results are
reported within each Stage, by group comparison. All peak voxel coordinates are reported
in X,Y,Z format.
Table 4: Voxel-Based Morphometry Results, Exploratory Analysis 2 (Gender)
Comparison Hemisphere Result p-value
All Females > All Males Right POP/PTR p=0.0030**
Left Parahippocampal
Gyrus
p=0.0264*
Bilateral Anterior Cingulate
Gyrus
p=0.0616ᵗ
Right Temporal Pole/
Orbitofrontal Cortex/
Insula
p=0.0954ᵗ
All Males > All Females Right Superior/Lateral
Cerebellum
p=0.0742ᵗ
Architecture Females > Architecture Males Left Anterior Cingulate
Gyrus
p=0.0366*
Architecture Males > Architecture Females Left Superior Cerebellum p=0.0512ᵗ
Music Females > Music Males Right Precentral Gyrus/POP p=0.0134*
Music Males > Music Females N/A
Architecture Females > Music Females N/A
Music Females > Architecture Females N/A
Architecture Males > Music Males Left Superior/Posterior
Cerebellum
p=0.0384*
Music Males > Architecture Males Left Anterior Cingulate
Gyrus
p=0.0918ᵗ
All data analyzed and statistics processed via FSL. Table shows only very statistically significant (**),
statistically significant (*), or statistical trend (ᵗ) results.
87
a. Stage 1
Stage 1 compared six groups: architecture females, architecture males, music
females, music males, and all females and all males. No statistically significant voxelwise
differences were found in this analysis. However, some statistical trends toward
significance were found: in architecture males when compared to music males in L
superior/posterior cerebellum (54, 25, 25; p=0.094); and in all females when compared to
all males in R POP/middle frontal gyrus (20, 74, 48; p=0.0742).
b. Stage 2
Stage 2 compared the same six groups, within each of the ten a priori defined
ROIs. Statistically significant voxelwise differences were found: in architecture males
when compared to music males in L superior/posterior cerebellum (54, 25, 25;
p=0.0384*); in architecture females when compared to architecture males in L anterior
cingulate gyrus (50, 82, 42; p=0.0366*); in music females when compared to music
males in R precentral gyrus/POP (18, 67, 40; p=0.0134*); and in all females when
compared to all males in L anterior parahippocampal gyrus (57, 61, 20; p=0.0264*), and
in R POP/PTR (20, 73, 48; p=0.003**).
Statistical trends toward significance were found: in music males when compared to
architecture males in L anterior cingulate gyrus (48, 74, 51; p=0.0918); in architecture
males when compared to architecture females in L superior cerebellum (53, 33, 27;
p=0.0512); in music females when compared to music males in L inferior POP (69, 70,
34; p=0.0598); in all females when compared to all males in bilateral anterior cingulate
88
gyrus/paracingulate gyrus (45, 87, 40; p=0.0616), and in R temporal pole/orbitofrontal
cortex/insula (23, 72, 28; p=0.0954); and in all males when compared to all females in R
superior/lateral cerebellum (18, 34, 21; p=0.0742).
ii. Discussion
Despite the statistical significance and interesting nature of these results, one must
take into consideration the fact that freshmen and seniors are lumped together for this
analysis. Even so, there are only 8 female and 8 male architecture majors, and 15 female
and 10 male music majors. Given the small numbers, pronounced individual differences
(outliers) may again be driving some of these results.
Statistically significant results and statistical trends were found in both directions
in the comparisons between all females and all males. Females were found to have
statistically significantly increased grey matter in right POP and PTR, and in left anterior
parahippocampal gyrus, when compared to males. Results in right POP and PTR
corroborate those of Good (et al, 2001), who reported larger grey matter volume in
inferior frontal gyrus, for females in a large sample of normal adult subjects. A similar
gender difference for the hippocampus was reported by Giedd (et al, 1996), who found
greater right than left laterality in the hippocampus for females in late adolescence (18
years). The authors attributed the increase in hippocampal volume in females to the
relatively higher number of estrogen receptors found in the primate hippocampus, as
compared to androgen receptors found in the amygdala (Giedd et al, 1996, 2008).
Though my results were in the opposite hemisphere, the reasoning – estrogen receptor
location – holds for both hemispheres.
89
In addition to the statistically significant POP, PTR, and parahippocampal gyrus
results, statistical trends toward increased grey matter were found for females in bilateral
anterior cingulate gyrus and across right temporal pole, orbitofrontal cortex, and insula.
The anterior cingulate gyrus has been linked in functional imaging studies to both
emotion-related and creative tasks: positive affect achieved by viewing beautiful stimuli
(Nadal et al, 2008), aesthetic appreciation (Chakravarty, 2012), emotion regulation
(reviewed in Vaidya et al, 2007, Mann et al, 2011), the implementation of executive
control during creative tasks (Howard-Jones et al, 2005), and mind-wandering (Mason et
al, 2007, reviewed in Craig, 2009). Results from the aforementioned studies were
bilateral. While both female architecture majors and female music majors are
theoretically involved in emotional and creative tasks on a regular basis, it is difficult to
extend this line of reasoning to gender differences, unless one is to argue that females are
more creative and/or more emotional – or better at regulating their emotions – than
males. Gender differences in anterior cingulate gyri, as evidenced by the extant literature,
are present in healthy adults, and offer a more likely explanation. As reviewed by Mann
(et al, 2011), women typically have larger cingulate gyri than men, and tend to have more
symmetrical cingulate gyri (men typically have a leftward cingulate asymmetry).
Regarding the final statistical result for all females when compared to all males, one must
note that the finding of increased grey matter in the vicinity of right temporal pole,
orbitofrontal cortex, and insula is at best a very weak statistical trend (p=0.0954).
Reasons for morphological gender differences in these right hemisphere regions are
unknown.
90
Statistical trends toward increased grey matter were found for males when
compared to females in right superior lateral cerebellum. Aside from its known role in
motor movement and musical training, which applies to both genders, larger cerebellar
grey matter volumes have been previously reported in men than in women (reviewed in
Lentini et al, 2012), supporting the current finding.
Possible gender differences were investigated within and between majors. Within
majors, differences were found between female architecture majors and male architecture
majors, and in female music majors when compared to male music majors. No
differences were found in male music majors when compared to female music majors.
Within architecture majors, females exhibited statistically significant larger grey
matter in left anterior cingulate gyrus. Given that, as reviewed by Mann (et al, 2011),
women tend to have both larger and more symmetrical cingulate gyri than men (who
typically have a leftward asymmetry), one would expect to see voxelwise differences in
female architecture majors when compared to male architecture majors primarily in the
right anterior cingulate gyrus. Interestingly, asymmetry was present but in the opposite
direction; there was leftward asymmetry,. The studies reviewed by Mann (et al, 2011)
looked at adults age 20 and higher; perhaps the younger freshman sample is biasing the
results. One also wonders why there were no statistically significant differences in left
hemisphere anterior cingulate in male architecture majors when compared to female
architecture majors. If this difference is a developmental issue, it might be possible that
males are developing more slowly than the females; such a difference in development
would yield larger cingulate results for females as compared to males, potentially
91
rightward and/or leftward, but not a leftward asymmetry in males as compared to females
– at least not until development is finished.
A statistical trend was found in male architecture majors when compared to
female architecture majors in left superior cerebellum. As with the finding of increased
left cerebellar grey matter in architecture freshmen when compared to seniors, discussed
previously, one cannot comment on the potential role of non-musical motor experiences
in this population, though a male-biased gender difference could be expected, given the
extant literature (reviewed in Lentini et al, 2012).
Female music majors were found to have larger grey matter in right lateral and
inferior precentral gyrus and superior POP when compared to male music majors. These
POP results are in a very similar location to those found when all females are compared
to all males, suggesting that the female music majors are driving those results. As I said
for those results, the right POP differences between female music majors and male music
majors can likely be attributed to gender differences in inferior frontal gyrus (Good et al,
2001).
Between majors, no differences were found between female architecture majors
and female music majors. Male music majors exhibited a weak statistical trend in left
anterior cingulate gyrus when compared to male architecture majors. As, according to
Mann (et al, 2011), men typically exhibit a leftward cingulate asymmetry, the results here
could suggest that the male architecture majors in this population have more symmetrical
cingulate gyri than the male music majors. Of course, there are only 10 male music
majors and 8 male architecture majors in this sample; it is, as always, possible to attribute
findings to pronounced individual differences.
92
Male architecture majors were found to have statistically significantly increased
grey matter in left superior/posterior cerebellum when compared to male music majors.
Musical training is clearly not the culprit here, although, as with other cerebellum results
in the present study, there could be other forms of motor experience at work. The lack of
cerebellum results in male music majors suggests that male and female music majors are
more similar, in terms of cerebellar grey matter, than male and female architecture
majors.
While some of these gender differences, particularly in inferior frontal gyrus and
cerebellum, align with the extant literature, some, such as those in anterior cingulate
gyrus, do not. It would be interesting to know if any of these latter results may be present
in a larger population of post-collegiate architects and musicians. With the exception of
the music senior group, gender is balanced evenly in this dataset. Thus, unlike the results
of the first exploratory analysis, on instrumentation, these gender effects should have
little effect on the planned comparisons by type of major and years of training discussed
earlier in this chapter.
The final exploratory analysis was undertaken on the architects with significant
musical training (AWM), the 8 architecture majors removed from the original
architecture group after descriptive statistical analysis of the performance background
questionnaire, detailed in chapter two.
93
3. Architects with Significant Musical Training
i. Results
Once again, VBM analyses were performed in two stages, summarized in Table 5.
Stage 1 made voxelwise comparisons across the whole brain between six groups: music
seniors, music freshmen, architecture seniors, architecture freshmen, and architecture
majors with significant musical training (AWM), at both the freshman and senior levels.
This analysis only pertains to the comparisons between music groups and AWM groups,
and architecture groups and AWM groups; the exploration was fueled by an interest in
whether or not the AWM groups could be considered a true middle group. Stage 2 made
voxelwise comparisons within each of the a priori defined ROIs, to improve statistical
power. Statistically significant and statistical trend results are reported within each Stage,
by group comparison. All peak voxel coordinates are reported in X,Y,Z format.
a. Stage 1
Stage 1 compared both AWM groups (freshmen and seniors) to both music
groups and both architecture groups. No statistically significant voxelwise differences or
statistical trends were found during this analysis.
b. Stage 2
Stage 2 compared the same six groups, within each of the ten a priori defined
ROIs. Statistically significant voxelwise differences were found: in music seniors
compared to AWM seniors in R cerebellum (37, 29, 11; p=0.0346*).
94
Table 5: Voxel-Based Morphometry Results, Exploratory Analysis 3 (Architects with Significant
Musical Experience (AWM))
Comparison Hemisphere Result p-value
All Music Majors > All AWM Left Postcentral Gyrus p=0.0734ᵗ
All AWM > All Music Majors N/A
All Architecture Majors > All AWM Right Postcentral Gyrus/
Supramarginal Gyrus
p=0.0878ᵗ
All AWM > All Architecture Majors N/A
Music Seniors > AWM Seniors Right Inferior/Medial
Cerebellum
p=0.0346*
Left Anterior Insula p=0.0538ᵗ
AWM Seniors > Music Seniors Left Posterior Fusiform
Gyrus/
Parahippocampal Gyrus
p=0.0518ᵗ
Right Planum Polare/
Heschl’s Gyrus
p=0.0874ᵗ
Music Freshmen > AWM Freshmen Left Precentral Gyrus/
Postcentral Gyrus
p=0.0782ᵗ
AWM Freshmen > Music Freshmen N/A
Architecture Seniors > AWM Seniors N/A
AWM Seniors > Architecture Seniors Left Parahippocampal Gyrus p=0.0834ᵗ
Architecture Freshmen > AWM Freshmen Bilateral Frontal Pole/
Orbitofrontal Cortex
p=0.0820ᵗ
AWM Freshmen > Architecture Freshmen N/A
AWM Seniors > AWM Freshmen N/A
AWM Freshmen > AWM Seniors Right Middle Cingulate
Gyrus
p=0.0624ᵗ
All data analyzed and statistics processed via FSL. Table shows only very statistically significant (**),
statistically significant (*), or statistical trend (ᵗ) results.
Statistical trends toward significance were found: in AWM freshmen when compared to
AWM seniors in R middle cingulate gyrus (42, 63, 58; p=0.0624); in all architecture
majors when compared to all AWM subjects in R postcentral gyrus/supramarginal gyrus
(27, 42, 61; p=0.0878); in architecture freshmen when compared to AWM freshmen in
bilateral orbitofrontal gyrus/PTR (18, 81, 32; p=0.082); in AWM seniors when compared
to architecture seniors in L anterior parahippocampal gyrus, partially in white matter (36,
52, 30; p=0.0834); in AWM seniors when compared to music seniors in L posterior
fusiform gyrus/anterior parahippocampal gyrus (64, 53, 28; p=0.0518), and in R planum
polare/HG (18, 61, 36; p=0.0874); in all music majors when compared to all AWM
95
subjects in L postcentral gyrus (66, 44, 68; p=0.0734); in music seniors when compared
to AWM seniors in L anterior insula (61, 71, 33; p=0.0538); and in music freshmen when
compared to AWM freshmen in L precentral gyrus/postcentral gyrus (59, 49, 66;
p=0.0782).
ii. Discussion
This exploratory analysis was intended to investigate whether or not the AWM
group could be considered a true middle group. Contrary to expectations, the AWM
group exhibited voxelwise results when compared to music majors in two particular ROIs
that were predicted to yield largest results for music majors: parahippocampal gyrus and
HG. Both of these results were statistical trends, found in AWM seniors when compared
to music seniors. While all results in this exploratory analysis are certainly interesting, it
is difficult to explain their existence given that there are so few subjects in each AWM
group (3 freshmen, 5 seniors), one of whom, an AWM senior, is the only left-handed
subject in the entire dataset. There are likely outliers driving all of the results, which can
only be clarified by individual subjects investigation, something that requires the use of
different methods of analysis.
E. GENERAL DISCUSSION AND CONCLUSION
No results were found in the following hypothesized regions of interest for music
students when compared to architecture students: precentral gyrus, postcentral gyrus,
inferior frontal gyrus, hippocampus, Heschl’s gyrus, cingulate gyrus, and cerebellum.
96
The lack of differences between music students and architecture students in inferior
frontal gyrus and HG could be due to differences among the instrument groups.
Regarding the other regions of interest, and as discussed in the previous chapter, it is
possible that four years of conservatory training were simply not enough for the students
in this sample to incur significant grey matter density increases in these areas.
Alternatively, the lack of differences could be linked to the end of the regular
developmental period, or to the relatively small sample size of the dataset, or these areas
simply might not be larger in musicians compared to non-musicians. It seems more
likely, however, that the results are either related to the control population (architecture
students) who may be more similar in brain morphology to the music students than had
been expected, or, on the other hand, to the method of analysis itself.
To conclude, the results generated by this voxel-based morphometry analysis are
intriguing, often unexpected, and somewhat challenging to explain. While statistically
significant voxelwise differences were found in the right cerebellum of music seniors
when compared to music freshmen, no other predicted results were found in the within
major, by year comparison, in either music students or architecture students.
Additionally, unexpected gender differences appeared, particularly in one comparison
within architecture majors: in left anterior cingulate gyrus of females when compared to
males. As VBM is by necessity a group analysis tool, it cannot take into account
significant individual differences among subjects.
97
i. The Study-Specific Template
In addition to the other shortcomings of VBM, the changing of subjects included
in the study-specific template appears to subtly shift the location and/or significance of
results. For VBM to work, an equal representation of each group population must be
present within the study-specific template. This means that if, for example, there are 6
piano majors, 9 strings majors, 10 voice majors, and 10 architecture seniors, 6 subjects
from each group must be included in the template. The 6 from each larger group must be
chosen carefully, in order to include all variations represented within that group. Within
the current study, these variations included gender, year in school, and instrument group.
For example, in the comparison between instrument groups, the template was built with
as close to the same number of men, women, freshmen, and seniors representing each
instrument group as possible. For the comparison between architecture majors and music
majors, the same number of men, women, freshmen, and seniors were chosen for the
template. Additional attention was given to choosing subjects from all three instrument
groups (piano, strings, and voice) for this template. If the numbers of these group
representatives within the study-specific template change, as they do when there are more
people in some groups, results can be affected. For example, parts of the first exploratory
analysis were conducted twice, to help validate the comparisons between instrument
groups. Figure 2a shows the results yielded for architecture majors when compared to
voice majors in the first run, with 6 subjects (the number of subjects in the piano major
group) from each group represented in the template. The groups in the first run were
piano majors, strings majors, voice majors, and architecture seniors. For the second run,
piano majors and strings majors were combined to form the instrumentalists group.
98
Figure 2a: Architecture Seniors > Voice Majors (first run), p=0.0582
Figure 2b: Architecture Seniors > Voice Majors (second run), p=0.0976
Figure 2b shows the results of the comparison between architecture majors and vocalists
(voice majors; the same comparison seen in Figure 2a) yielded in the second run, with 10
subjects from each group (the number of subjects in each of the architecture seniors and
99
the voice majors groups) represented in the template. The addition of 4 subjects to the
template slightly shifted the location of the results (see the atlas descriptions), and
changed the significance from a strong statistical trend in Figure 2a (p=0.0582) to barely
a statistical trend (p=0.0976) in Figure 2b. Clearly, either a shift in region or a shift in
significance alters the interpretation of the data; the existence of this issue calls into
question the reliability of VBM results, particularly when comparing groups of different
sizes, as the template must include equal numbers of subject brains from each represented
group, and has documented difficulty accounting for non-normality within groups
(Scarpazza et al, 2013).
Had I included the same brains in the template for all analyses, this issue would
not have been discovered. In order to perform all the analyses previously described using
one template, the groups included within the study-specific template would have had to
be divided as follows: architecture freshmen (n=6), architecture seniors (n=10), AWM
subjects (n=8), piano major freshmen (n=3), piano major seniors (n=3), strings major
freshmen (n=5), strings major seniors (n=5), voice major freshmen (n=2), and voice
major seniors (n=7). A comparison between architecture freshmen and music freshmen
would have involved the comparison between the former group with a combination of
piano, strings, and voice major freshmen. VBM is capable of doing this. However, as the
study-specific template must be built from the same number of subjects in each group,
the above distribution of subjects would have required that all groups have only 2
subjects, the number of voice major freshmen, in the template. Two subjects would
certainly not be truly representative of each group, and would decrease normality,
increasing the rate of false positives in results. I thus decided to build different study-
100
specific templates for the planned and exploratory analyses. The first included
architecture freshmen, architecture seniors, AWM subjects, music freshmen, and music
seniors. The second included piano majors, strings majors, voice majors, and architecture
seniors. The third included instrumentalists, vocalists, and architecture seniors, yielding
the redundant comparison of vocalists and architecture seniors (Figure 2a-b), and alerting
me to this issue with the study-specific template. The fourth included architecture
females, architecture males, music females, and music males. The fifth and final template
included architecture freshmen, architecture seniors, AWM freshmen, AWM seniors,
music freshmen, and music seniors. With only 3 subjects per group, results using this
final template are indeed the most questionable of the entire VBM process.
The next logical step in analysis of this dataset is to use a more classical approach
on the data; to use methods that, while more manual, yield results that can be parsed by
individual subjects and do not require group comparison to a study-specific template.
Such an approach is detailed in the following chapter.
101
CHAPTER 4
A. BACKGROUND
Studies of experience-dependent musical structural neuroplasticity that employ
manual tracing methods are typically limited to just a few – if not only one – region(s) of
interest. This is due to the time-consuming nature of these methods, which often involve
an additional tracer on at least a subset of the dataset, in order to achieve reliable results.
Automated methods, such as VBM, offer speed, but often sacrifice accuracy in results. A
semi-automated method using the software package BrainSuite proposes to walk the
middle line between automated structural analysis and manual tracing (Shattuck et al,
2009). While manual editing is still required to achieve good brain extraction from the
skull and other tissues, to fix minor issues at other steps of the process, and to mark sulci
on the 3D brain generated by the software, the program uses these markings to constrain
the registration of an atlas brain to each subject’s brain. This process yields relatively
clean registration in native space, allowing for exploration of individual differences
between subjects. Additionally, BrainSuite delivers volumetric data – total volume, grey
matter, white matter, and cortical thickness - for all brain regions, not just the few chosen
a priori for study. The process is not without issues, which will be discussed in more
detail below. However, as BrainSuite is updated and improved regularly, the method
holds much promise for structural neuroimaging research.
In addition to the volumetric measurements listed above, diffusion tensor imaging
(DTI) data for each subject was co-registered to structural volumes, and diffusion
measures were produced for all regions. Fractional anisotropy (FA) was selected prior to
analysis as the measure of choice, given that it has been used in the extant music
102
literature in relation to white matter volume and as a measurement of white matter
integrity. As this dataset contains both FA and WM volume data for the same regions of
interest, I conducted correlational analyses between the two measurements. FA and WM
(both absolute and relative volumes) only correlated in the following regions: left POP,
right Heschl’s gyrus, right fusiform gyrus, and bilateral insula. These findings suggest
that FA is not reliably correlated with WM volume. FA values represent the degree of
anisotropy and directionality of the diffusion tensor. With a value of 0, one can expect
that unrestricted, spherical (isotropic) diffusion of water within the brain is occurring. A
value of 1 would represent restricted/hindered (anisotropic) diffusion, traveling in only
one direction. Diffusion can be hindered by many factors, including elements of
microstructure that one might expect, such as the myelin sheath, axon diameter, fiber
density, and cell membranes. Other factors affecting diffusion of water molecules can
include, but are not limited to, temperature, viscosity, and other large molecules (Jones et
al, 2013). FA values fall accordingly between 0-1, with all of the above factors, and
more, contributing to anisotropy (Alba-Ferrara & de Erausquin, 2013, Jones et al, 2013,
Soares et al, 2013). FA values can be decreased by crossing fibers, which is of great
concern, given that some studies have found that up to 90% of white matter voxels
contain at least some crossing fibers (Alba-Ferrara & de Erausquin, 2013). Additionally,
FA values can be decreased when close proximity to neighboring grey matter voxels
results in partial volume effects (Leow et al, 2009), or when water molecules move in
any direction other than perpendicular to the applied gradient axis (Jones et al, 2013). FA
values should thus be used, not as a proxy for white matter volume or as an indication of
white matter integrity, but as a general indication of white matter microstructure, affected
103
by a variety of factors. For example, lower FA values in one group as compared to
another could very well indicate a loss in myelination; they could also indicate a lower
packing density, increased cell membrane permeability, or larger axon diameters (Jones
et al, 2013). It is now generally recommended to use several different diffusion measures
(some combination of FA, mean diffusivity, radial diffusivity, and/or axial diffusivity) to
paint a clearer picture of tissue microstructure (Alexander et al, 2007). As a companion to
WM volume measurements, as in this study, FA values deliver some insight into the
existence of differences in white matter microstructure between groups. These are healthy
populations, and as such, inferences about white matter integrity really cannot be made.
There is some suggestion that FA values change prior to WM volumes; thus, a decrease
in FA could precede a decrease in WM volume, as can be seen in older populations
(Hugenschmidt et al, 2008). If this is the case, it is possible that an increase in FA values
could precede an increase in WM volume; this would help explain some FA – but not
WM – results seen in the present study and presented below, particularly in hippocampus,
precentral gyrus, and postcentral gyrus. It would not explain the lack of FA results in
areas where WM results are seen, however. More likely is the explanation that these
results exist due to microstructural differences in these populations that remain
undetectable without the addition of further diffusion measures to the analysis, and/or the
use of additional modeling procedures. It should be mentioned that, although
hypothesized to be regions of interest in this analysis, cerebellum, corpus callosum, and
the posterior limb of the internal capsule were not investigated. This decision was based
on BrainSuite limitations (at the time), and in the interest of project completion. Recent
improvements to the software have delivered the ability to look at the cerebellum, and the
104
use of additional tools should allow for investigation of both corpus callosum and internal
capsule. These three regions of interest, along with the use of additional diffusion
measures, can be considered for future directions for this work.
B. METHODS
i. Subjects
The analyses in this chapter were performed in the same subject population
described in both chapters 2 and 3. Within the architecture majors, there were 6 freshmen
and 10 seniors. Within the music majors, there were 10 freshmen and 15 seniors. Among
these, there were 50% males in the architecture majors, and 40% males in the music
majors.
ii. Image Acquisition
High-resolution anatomical MRI and Diffusion MRI scans were obtained in all
subjects following the protocol described in chapter 3. Semi-automated structural
analyses were performed on all MPRAGE and diffusion data using BrainSuite software.
iii. Analysis
1. Structural MRI data
BrainSuite (Shattuck & Leahy, 2002) includes a six stage cortical modeling sequence for
analyzing T1-weighted MR volumes. The following is a brief description of the method
as used in this study: First the brain was extracted from the surrounding skull and scalp
105
tissues using a combination of edge detection and mathematical morphology, an image
processing technique that fills holes in brain masks, reduces and/or increases the size of
masks, and makes the masks more smooth, among other things. Parameters for edge
detection were adjusted to improve each individual mask. Manual editing was next used
on each coronal slice to ensure the best extraction. Second, MRI intensities were
corrected for non-uniformity of pixel values, produced by MRI scanner bias-field
inhomogeneities. Third, each voxel in the corrected image was labeled according to tissue
type using a statistical classifier. Fourth, a standard atlas with associated structure labels
was aligned to the subject volume, providing a label for cerebellum, cerebrum, and
brainstem. These labels were combined with the tissue classification to automatically
identify the grey/white boundary, to fill the ventricular spaces (so as not to have them
misclassified as white or grey matter), and to remove the brainstem and cerebellum. The
cerebrum mask yielded during this step was manually edited for all subjects, to ensure
inclusion of all cerebral, and exclusion of all brainstem and cerebellar, tissue. Fifth, a
surface was generated automatically corresponding to the grey/white separation (inner
cortical mask) in the cerebrum, referred to as the inner cortical surface. The inner cortical
mask was also manually edited to improve the subsequent surface generation.
Tessellation of the grey/white matter mask, used to produce the inner cortical surface,
often produces a surface with topological handles, meaning that the surface will include
artifacts, holes or bridges across hemispheres. These are not anatomically correct, and are
due to finite voxel size. In order to avoid these errors, these handles were identified and
removed from the binary volume automatically using a graph-based approach, prior to
tessellation (Shattuck & Leahy, 2002). Sixth, a tessellated isosurface of the resulting
106
inner cortical mask was extracted to produce a genus zero (inner cortical) surface based
on the registered atlas labels; the target brain was subsequently split into two cortical
hemisphere surfaces (Shattuck et al, 2009, Joshi et al, 2010). Three types of surfaces
were generated (Figure 1): the inner cortical surface, mid-cortical surface, and pial
surface. The inner cortical surface, the original genus zero surface, is generated from the
grey matter/white matter divide, throughout the entire brain. The pial surface generation
module takes as input the initial grey matter/white matter mask and the surface defined
by its boundary. It also uses a set of tissue fraction values that define how much GM,
WM and CSF are contained within each voxel of the brain image. Using an iterative
process, each vertex of the WM surface is moved under the influence of several forces.
The first pushes the vertex outward along the surface normal; the other forces try to
maintain the smoothness of the initial surface. The movement of each vertex stops when
it moves into a location with a significant CSF component or when its motion would
cause the surface to self-intersect. The result is a one-to-one map between the points on
the inner cortical surface model and the pial surface model, which can provide a direct
estimate of the cortical thickness. Finally, the mid-cortical surface is generated from the
halfway point between the grey matter/white matter divide and the generated pial surface.
In order to improve co-registration of atlas labels to each individual subject’s
brain, 27 sulci were identified and marked in each hemisphere, for each subject, on the
mid-cortical surface. The reason to identify and trace the sulci on this surface is that it is
easier to delineate the sulci on this more open surface, which allows a better view of the
sulci without making them so wide as to make their precise identification more difficult
107
Figure 1: The three surfaces generated by BrainSuite, marked sulci, and the final output, a labeled pial
surface. All images are in right hemisphere. A: The genus zero (inner cortical) surface. B: The mid-cortical
surface. C: The pial surface. D: Sulci marked on the mid-cortical surface. E: The final labeled pial surface.
108
(which is the case with the inner surface). See figure 1. Following sulcal marking, a
specific script (svreg) was used to generate, in the target brain, all individual ROIs
constrained by the labeled sulci (Figure 1). For more details see the description of Svreg
in Joshi et al, 2012. This process yields well-labeled surfaces and volumes. The use of the
initial atlas produces one notable exception in the left hemisphere in the precentral and
postcentral gyri, due to a morphological particularity of the atlas brain. A new atlas, using
a brain without such deviation from the standard anatomy, has been recently developed.
Not all regions in the new atlas (BCI-DNI atlas) were available at the time of this study;
while all precentral and postcentral values were obtained on the basis of the BCI-DNI
atlas, all other ROI values were obtained on the basis of the initial atlas (Colin atlas).
Subcortical labeling is not yet completely reliable. The amygdala, for example, can often
be misplaced and therefore is not included in this analysis. There were additional
problems with registration in specific frontal and occipital regions due to an
unsatisfactorily corrected issue of the bias-field’s inhomogeneity. Affected regions
relevant to this study not used for this reason were orbitofrontal cortex and temporal pole.
Recently, a tool to enable better manual correction has been introduced; unfortunately,
this improvement arrived too late to be used in the current study, but will be of great use
to future studies using this method.
2. Diffusion data
For each individual subject, the diffusion MRI data was co-registered with the structural
MRI data (MPRAGE) using the constrained non-rigid registration technique described in
109
Bhusan et al, 2012. This registration technique is based on mutual information from the
two data sets, and uses MPRAGE and T2 weighted b=0 images to estimate susceptibility-
induced distortion in diffusion data acquired with EPI pulse sequences. The deformation
was constrained in two ways: along the phase-encoding direction of the diffusion images,
and by adding spatial regularization, which smoothed the deformation. During
registration, intensity of diffusion images was modified to account for localized MRI
signal accumulation and dispersion. The co-registered diffusion images were warped to
the anatomical template (MPRAGE). Next, diffusion tensors were estimated at each
voxel following the protocol established in Kim et al, 2009. Eigenvalues and
eigenvectors were used to generate diffusion measures, including fractional anisotropy,
axial diffusivity, radial diffusivity, and mean diffusivity. Following this, BrainSuite labels
from the structural MRI analysis were applied, so as to calculate diffusion values
(including FA) for each region of interest.
3. Statistical Analysis
Statistical comparisons were handled in SPSS. Five values were analyzed per
region of interest: total volume (TV), grey matter volume (GM), white matter volume
(WM), cortical thickness (CT), and fractional anisotropy (FA). To account for total brain
size differences among subjects values relative to brain size were generated for total
volume, grey matter volume, and white matter volume for each region of interest. All TV,
GM, and WM results discussed in this chapter are based on these relative values, not the
absolute values. Independent samples t-tests (simple statistical tests used to compare
mean values between groups, in order to determine whether or not there exists a
110
statistically significant relationship between those two groups for specific variables),
were used to compare means of each value (TV, GM, WM, CT, and FA), for all relative
group comparisons (music majors/architecture majors, freshmen/seniors, etc.), within
each a priori defined ROI. Boxplot inspections yielded outlier information for each t-test;
each t-test was performed with and without outliers. The Holm-Bonferroni adjustment
was applied to all p-values to correct for multiple comparisons within each region of
interest. Statistically significant results are presented in Tables 1-3. As discussed in
chapter 2, with sample sizes of less than 18 subjects per group (the number yielded by my
power analysis of the extant literature), statistical trends were more expected than
statistically significant results in the current dataset; statistical trends are thus also
included in all tables. A record of all comparisons can be found in Tables 5-9, at the end
of this chapter. The list of regions of interest analyzed, guided by hypotheses and VBM
results, includes: superior frontal gyrus, middle frontal gyrus, inferior frontal gyrus,
Broca’s area, POP, PTR, pars orbitalis, precentral gyrus, postcentral gyrus, cingulate
gyrus, supramarginal gyrus, angular gyrus, fusiform gyrus, superior temporal gyrus,
Heschl’s gyrus, parahippocampal gyrus, hippocampus, and insula.
Correlational analyses were performed for several questionnaire variables and
specific regions of interest. Scatterplots were generated for each desired comparison.
Spearman’s Rho – chosen due to negatively-skewed monotonic scatterplots - was used to
assess correlational strength and direction between eleven regions of interest and age of
onset of musical training (for both the first instrument and each subject’s primary
instrument), and intensity of practice, as determined by average hours spent practicing the
primary instrument each day. The eleven regions of interest analyzed, guided by
111
hypotheses and results generated from previous analyses, were: Broca’s area, POP, PTR,
precentral gyrus, postcentral gyrus, cingulate gyrus, superior temporal gyrus, Heschl’s
gyrus, parahippocampal gyrus, hippocampus, and insula.
C. PROPOSED/ORIGINAL ANALYSES
i. Results
BrainSuite analysis, unlike VBM analysis, benefits from the fact that volumetric
data is produced at the same time for all brain regions delineated by the protocol
followed. This renders the 2-stage process of statistical analysis performed in VBM –
whole brain analysis followed by individual region of interest analysis - unnecessary in
BrainSuite. Based on my hypotheses and some results generated with the use of VBM, I
chose to investigate possible differences within the following a priori defined ROIs,
separately for each hemisphere: superior frontal gyrus, middle frontal gyrus, POP, PTR,
pars orbitalis, Broca’s area (POP + PTR), inferior frontal gyrus (POP, PTR, and pars
orbitalis), precentral gyrus, postcentral gyrus, cingulate gyrus, supramarginal gyrus,
angular gyrus, superior temporal gyrus, Heschl’s gyrus, fusiform gyrus, parahippocampal
gyrus, hippocampus, and insula.
Six groups were compared for the planned analysis: all music majors, all
architecture majors, music seniors, music freshmen, architecture seniors, and architecture
freshmen. All group analysis results were run with and without outliers, to investigate
possible influences of individual differences. Results are presented with measurement
type (relative total volume (TV), relative grey matter volume (GM), relative white matter
volume (WM), cortical thickness (CT), or fractional anisotropy (FA)) first, followed by
112
corrected p-value. All statistically significant and statistical trends results for this analysis
are presented in Table 1; all results for this analysis are presented in Tables 6-7, at the
end of this chapter.
Statistically significant volumetric and/or FA results were found: in all music
majors compared to all architecture majors in left Broca’s area (TV: p=0.0222*, GM:
p=0.0198*) and
left inferior frontal gyrus (TV: p=0.0327*, GM: p=0.0408*); in all architecture majors
compared to all music majors in right hippocampus (FA: p=0.0160*); in architecture
seniors compared to music seniors in right hippocampus (FA: p=0.0497*); in music
freshmen compared to architecture freshmen in left PTR (TV: p=0.0396*, GM:
p=0.0467*), left Broca’s area (TV: 0.0169*, GM: p=0.0181*, WM: p=0.0232*), and left
inferior frontal gyrus (TV: p=0.0106*, GM: p=0.0176*, WM: p=0.0137*); in music
seniors compared to music freshmen in left postcentral gyrus (TV: p=0.0074**, GM:
p=0.0113*, WM: p=0.0070**); in architecture seniors compared to architecture freshmen
in right Broca’s area (WM: 0.0490*); and in architecture freshmen compared to
architecture seniors in left precentral gyrus (TV: p=0.0229*, GM: p=0.0027**).
Statistical trends toward significance were found: in all music majors when
compared to all architecture majors in left POP (TV: p=0.0608ᵗ, GM: p=0.0804ᵗ, WM:
p=0.0670ᵗ) and left Broca’s area (WM: p=0.0759ᵗ); in music freshmen compared to
architecture freshmen in left PTR (WM: p=0.0783ᵗ); and in architecture seniors compared
to architecture freshmen in right inferior frontal gyrus (WM: p=0.0645ᵗ).
With the exception of one result, all results either stayed the same or strengthened
in statistical significance with the removal of outliers. The exception was left inferior
113
frontal gyrus, in all music majors compared to all architecture majors. Removal of two
outliers, one in each group, decreased the result to less statistically significant in TV
(p=0.0436*), and to a weak statistical trend in GM (p=0.0990ᵗ). Two different outliers
were present within the music group
in the same comparison, in the WM measurement. When removed, left inferior frontal
gyrus WM reached a weak statistical trend (p=0.0920ᵗ).
ii. Discussion
No predicted statistically significant results or statistical trends were found in the
comparison between architecture majors and music majors, and within architecture major
and music major groups depending on year in school (either freshman or senior), in the
following regions of interest: precentral gyrus, hippocampus, Heschl’s gyrus, cingulate
gyrus, or insula. Cerebellum, orbitofrontal cortex, corpus callosum, and internal capsule
were not included in the BrainSuite analysis, for the reasons explained earlier.
Left Broca’s area of the inferior frontal gyrus appears as a robust finding in the
comparison between music majors and architecture majors. While increased TV, GM,
and WM volume in PTR only appear at the freshman level, increased TV, GM, and WM
volume in POP appear at the overall comparison level (all music majors compared to all
architecture majors), suggesting that the senior population is also contributing to the
statistically significant volumetric differences in Broca’s area, again in all three
volumetric measurements. As discussed in chapter 2, increases in grey matter density
and/or grey matter volume were predicted in inferior frontal gyrus, particularly on the
left, for music majors compared to architecture majors. This prediction was made due to
the extant literature, reviewed in chapter 1 (pg. 33), showing both structural and
114
Table 1: BrainSuite Results, Planned Analysis (Major and Year)
Comparison Hemi Type Result p-value
Music Majors > Architecture Majors Left TV POP 0.0608ᵗ
Left GM POP 0.0804ᵗ
Left WM POP 0.0670ᵗ
Left TV Broca’s Area 0.0222*
Left GM Broca’s Area 0.0198*
Left WM Broca’s Area 0.0759ᵗ
Left TV Inferior Frontal
Gyrus
0.0327*
Left GM Inferior Frontal
Gyrus
0.0408*
Architecture Majors > Music Majors Right FA Hippocampus 0.0160*
Music Seniors > Architecture Seniors N/A
Architecture Seniors > Music Seniors Right FA Hippocampus 0.0497*
Music Freshmen > Architecture Freshmen Left TV PTR 0.0396*
Left GM PTR 0.0467*
Left WM PTR 0.0783ᵗ
Left TV Broca’s Area 0.0169*
Left GM Broca’s Area 0.0181*
Left WM Broca’s Area 0.0232*
Left TV Inferior Frontal
Gyrus
0.0106*
Left GM Inferior Frontal
Gyrus
0.0176*
Left WM Inferior Frontal
Gyrus
0.0137*
Architecture Freshmen > Music Freshmen N/A
Music Seniors > Music Freshmen Left TV Postcentral Gyrus 0.0074**
Left GM Postcentral Gyrus 0.0113*
Left WM Postcentral Gyrus 0.0070**
Music Freshmen > Music Seniors N/A
Architecture Seniors > Architecture Freshmen Right WM Broca’s Area 0.0490*
Right WM Inferior Frontal
Gyrus
0.0645ᵗ
Architecture Freshmen > Architecture Seniors Left TV Precentral Gyrus 0.0229*
Left GM Precentral Gyrus 0.0027**
All data analyzed in BrainSuite and statistics processed via SPSS. Table shows only very statistically
significant (**), statistically significant (*), or statistical trend (ᵗ) results. All p-values are corrected for
multiple comparisons using the Holm-Bonferroni adjustment.
functional results in left inferior frontal gyrus in a variety of musical populations, both
professional and amateur, when compared to non-musicians. Functional usage of the left
inferior frontal gyrus has been linked to the processing of the syntax and semantics of
115
music (Sergent et al, 1992, Platel et al, 1997, Maess et al, 2001, Parsons, 2001, Koelsch
et al, 2005), and to music production (Clos et al, 2013). Higher WM volume in this
region could also be indicative of increased functional usage, and thus, increased
communication with other brain regions. No FA increases were found in IFG for this
comparison, again suggesting that FA may not be a reliable proxy for WM volume.
Architecture majors exhibited increased FA in right hippocampus, when
compared to music majors. This result appears in architecture seniors when compared to
music seniors, but not in the same comparison between majors at the freshman level,
leading one to suspect that the architecture seniors are driving the FA increase in right
hippocampus. There were no significant volumetric results in hippocampus. Without
further diffusion parameters, it is difficult to interpret what an increase in FA in right
hippocampus might mean for architecture majors compared to music majors.
When comparing freshmen and seniors within architecture majors or music
majors, we see that music seniors had increased TV, GM, and WM volumes in left
postcentral gyrus when compared to music freshmen. Postcentral gyrus, somatosensory
cortex, was predicted to be larger in both music majors compared to architecture majors
and in music seniors compared to music freshmen. These results compare well with those
reviewed in chapter 1 (pg. 30); multiple studies have found increased grey matter density
in musicians when compared to non-musicians, from childhood through adulthood.
Architecture results in relation to the number of years of study are less easy to
explain. Architecture seniors exhibited increased WM volume in right Broca’s area (and
inferior frontal gyrus, clearly driven by the Broca’s area result), when compared to
architecture freshmen, a difference not seen in the comparison of music seniors with
116
music freshmen. The right homologue to classical Broca’s area has been linked to
aesthetic judgment, demonstrated during a task involving judgment of beauty of
geometric shapes (Jacobsen et al, 2010). Continued involvement of right inferior frontal
gyrus for visual aesthetic judgment could yield increased WM matter volume; if so, one
would expect to see these increases in architecture seniors as compared to freshmen, as
we do in the current study. Aesthetic judgment is of particular importance to all artists
honing their skills during conservatory. Perhaps, given that processing of components of
musical perception occurs in left inferior frontal gyrus, aesthetic judgment of auditory
stimuli involves the left hemisphere as well. If aesthetic judgment of auditory stimuli
primarily involves left Broca’s area, rather than its right homologue, the lack of right
inferior frontal gyrus results in music majors might make sense.
Architecture freshmen exhibited increased TV and GM in left precentral gyrus as
compared to architecture seniors. Left precentral gyrus has been implicated in a left
hemisphere circuit used by creative artists and scientists in a verbal creativity task
(Andreasen et al, 2012). However, as precentral gyrus is primary motor cortex, it is likely
that the reason for its involvement in this task has more to do with the vocal production
aspect of the verbal task, rather than with creativity. This left precentral gyrus result also
appeared in the VBM analysis presented in chapter 3. While it is possible that these
architecture freshmen have had increased need of left precentral gyrus for unknown
motor reasons, it is also possible that, as discussed in chapter 3, these results could be
driven by development itself, given that grey matter density declines within the frontal
lobe through young adulthood (Gogtay et al, 2004, Thompson et al 2005, Whitford et al,
2007). If development is the reason for these results in architecture majors, one might not
117
expect to see increased precentral gyrus in music freshmen compared to music seniors,
given the high amount of pre-conservatory musical training, which will likely have
already altered the morphology of precentral gyrus. It is thus possible that increased
musical training during conservatory maintains, or slightly increases, the level of grey
matter volume in precentral gyrus in musicians, while volumetric decreases are seen
throughout young adulthood in non-musicians, including architects. This would explain
the results in architecture freshmen compared to architecture seniors as well as the lack of
results in music seniors compared to music freshmen.
As in chapter 3, several exploratory analyses were conducted to help clarify
possible difference between subgroups within the music major population, and any
differences between the two main majors groups and the possible middle group,
comprised of architecture majors with significant musical training (AWM).
D. EXPLORATORY ANALYSES
1. Instrumentation
i. Results
Five groups were compared for this exploratory analysis: piano majors, strings
majors, voice majors, instrumentalists (both piano and strings majors, compared to
vocalists (voice majors) and architecture seniors), and architecture seniors, chosen to
represent the architecture population. All group analysis results were conducted with and
without outliers, to investigate possible influences of individual differences. Results are
presented with measurement type (total volume (TV), grey matter volume (GM), white
matter volume (WM), cortical thickness (CT), or fractional anisotropy (FA)) first,
followed by corrected p-value. The results are presented as before with all statistically
118
significant and statistical trends results for this analysis in Table 2; all results for this
analysis can be found at the end of this chapter (Tables 8-9).
Statistically significant volumetric and/or FA results were found: in piano majors
compared to voice majors in right postcentral gyrus (FA: p=0.0141*); and in architecture
majors compared to vocalists in right hippocampus (FA: p=0.0171*).
Statistical trends toward significance were found: in piano majors compared to
strings majors in left precentral gyrus (GM: 0.0716ᵗ); in piano majors compared to voice
majors right precentral gyrus (FA: p=0.0585ᵗ) and right postcentral gyrus (FA:
p=0.0957ᵗ); and in instrumentalists compared to architecture majors in left POP (TV:
p=0.0663ᵗ, GM: p=0.0986ᵗ, WM: p=0.0652ᵗ), left Broca’s area (TV: p=0.0959ᵗ, GM:
p=0.0964ᵗ), and left inferior frontal gyrus (TV: p=0.0766ᵗ, GM: p=0.0882ᵗ).
With the exception of one result, all results either stayed the same or improved in
statistical significance with the removal of outliers. The exception was left postcentral
gyrus, in piano majors compared to voice majors. Removal of one voice major outlier
decreased the result in FA to below the threshold for a statistical trend (p=0.16). One
piano major outlier was decreasing significance in left precentral gyrus in the comparison
between piano majors and strings majors. When removed, left precentral gyrus GM
became a stronger statistical trend (p=0.0572ᵗ), and TV reached the level of statistical
significance (p=0.0250*).
119
Table 2: BrainSuite Results, Exploratory Analysis 1 (Instrumentation)
Comparison Hemi Type Result p-value
Instrumentalists > Vocalists N/A
Vocalists > Instrumentalists N/A
Piano Majors > Strings Majors Left GM Precentral Gyrus 0.0716ᵗ
Strings Majors > Piano Majors N/A
Piano Majors > Voice Majors Right FA Precentral Gyrus 0.0585ᵗ
Right FA Postcentral Gyrus 0.0141*
Left FA Postcentral Gyrus 0.0957ᵗ
Voice Majors > Piano Majors N/A
Strings Majors > Voice Majors N/A
Voice Majors > Strings Majors N/A
Instrumentalists > Architecture Majors Left TV POP 0.0663ᵗ
Left GM POP 0.0986ᵗ
Left WM POP 0.0652ᵗ
Left TV Broca’s Area 0.0959ᵗ
Left GM Broca’s Area 0.0964ᵗ
Left TV Inferior Frontal
Gyrus
0.0766ᵗ
Left GM Inferior Frontal
Gyrus
0.0882ᵗ
Architecture Majors > Instrumentalists N/A
Vocalists > Architecture Majors N/A
Architecture Majors > Vocalists Right FA Hippocampus 0.0171*
All data analyzed in BrainSuite and statistics processed via SPSS. Table shows only very statistically
significant (**), statistically significant (*), or statistical trend (ᵗ) results. All p-values are corrected for
multiple comparisons using the Holm-Bonferroni adjustment. Instrumentalists group is comprised of both
freshman and senior piano majors and strings majors; Vocalists group is comprised of both freshman and
senior voice majors. Architecture Majors group is comprised of senior architecture majors, to keep similar
group sizes. Piano, Strings, and Voice major groups include both freshmen and seniors.
ii. Discussion
Overall, the regions of interest revealed to be different between instrument groups
might be expected, given that they are motor-related regions. In the group of architecture
majors, the results in left inferior frontal gyrus in instrumentalists compared to
architecture majors clearly reflect the results found in the planned analysis for all music
majors compared to all architecture majors. While music freshmen seem to be driving the
PTR aspect of the latter result, discussed above, the current analysis makes it clear that
120
instrumentalists are driving the POP aspect of the overall music major result in Broca’s
area.
Piano majors exhibited larger GM volume in left precentral gyrus when compared
to strings majors. While the result is a statistical trend, it is nonetheless encouraging,
given the size of the groups. Increased left precentral gyrus volume for pianists would be
expected when compared to strings players, given the results reported by Bangert &
Schlaug (2006), and the manual requirements of keyboard and strings instruments, as
discussed in chapter 1. Despite the intense usage of left hand required for playing a
stringed instrument, increased right precentral gyrus volume might not be expected in
strings players compared to pianists, given that pianists also make regular and intense use
of left hand. Strings players likely make less intense use of right hand than pianists, given
that this hand is typically used only to bow and/or pluck strings; pianists need dexterous
use of both hands to play the piano well. The leftward asymmetry of hand usage by
strings players was given as the reason for the rightward asymmetry found by Bangert &
Schlaug; this decreased usage of right hand in comparison to the relative symmetry of
hand usage in pianists could lead to the increased left precentral GM result found here.
It appears to be the vocalists that are driving the right hippocampus FA results in
the comparison of architecture majors with music majors in the planned analysis
described above. In the current set of comparisons, the architecture majors exhibit
increased FA in right hippocampus compared to the vocalists. Perhaps a general
difference in white matter microstructure in right hippocampus in vocalists, and not in
instrumentalists, is the cause, although reasons for this are unclear.
121
FA increases for piano majors compared to voice majors in right precentral gyrus,
and bilateral postcentral gyrus, are also difficult to interpret. Any interpretation based on
hand usage
versus vocal cord usage would require further parcellation of both precentral and
postcentral gyri. Without this, and more diffusion parameters, all that can be said is that
these differences in FA indicate that these piano majors exhibit some combination of
microstructural differences in white matter when compared to the voice majors. In
general, the large number of voice majors within the music major population could very
well be influencing the overall results between music majors and architecture majors,
particularly in motor-related regions.
2. Architects with Significant Musical Training
i. Results
Three groups were compared for this exploratory analysis: all music majors, all
architecture majors, and all architecture majors with significant musical training (AWM).
All group analysis results were run with and without outliers, to investigate possible
influences of individual differences. Results are presented as before (see Tables 3 and
10). Statistically significant volumetric and/or FA results were found: in all AWM
compared to all music majors in right Heschl’s gyrus (CT: p=0.0001**); in all
architecture majors compared to all AWM in right Heschl’s gyrus (WM: p=0.0275*),
right parahippocampal gyrus (CT: p=0.0312*), and right hippocampus (GM: p=0.0288*,
FA: p=0.0156*); and in all AWM compared to all architecture majors in right Heschl’s
gyrus (CT: p=0.0037**)
122
Statistical trends toward significance were found: in all music majors compared to
all AWM in right pars orbitalis (TV: p=0.0863ᵗ, GM: p=0.0915ᵗ), and left angular gyrus
(TV: p=0.0838ᵗ, GM: p=0.0509ᵗ); and in all AWM compared to all architecture majors in
left POP (TV: p=0.0922ᵗ, GM: p=0.0830ᵗ).
Prior to processing any statistical t-tests, a standard boxplot inspection of each
comparison was performed. Boxplot inspection reveals which subjects fall outside a
standard deviation from the mean value of a given comparison; it is thus used to reveal
any outlier(s). This boxplot inspection process was repeated for each comparison;
depending on the ROI considered and the comparison in question, different subjects may
be found to be outliers. T-tests for each comparison were then performed with and
without outliers, the hope being that removal of outliers either would not alter the results
or, on the other hand, might have increased the significance of results. However, multiple
results did become less significant with the removal of outliers (Table 4). Removal of one
music major outlier in right pars orbitalis decreased the result in GM in music majors
compared to AWM so as not even to be considered a statistical trend (p=0.1600), and
removal of two music major outliers in TV decreased the result to an even lower
statistical level (p=0.2130). The removal of one music major outlier also decreased the
left angular gyrus result for GM to a weaker statistical trend (p=0.0850ᵗ). Removal of
four outliers (two each in the music major and AWM groups) reduced the statistical
significance of the CT result in right Heschl’s gyrus from p=0.0001** to p=0.0050**,
(still a very strong result). Also in right Heschl’s gyrus, removal of three outliers (one
architecture major and two AWM subjects), reduced the statistical significance of CT in
123
Table 3: BrainSuite Results, Exploratory Analysis 2 (Architects with Significant Musical
Experience (AWM))
Comparison Hemi Type Result p-value
All Music Majors > All AWM Right TV Pars Orbitalis 0.0863ᵗ
Right GM Pars Orbitalis 0.0915ᵗ
Left TV Angular Gyrus 0.0838ᵗ
Left GM Angular Gyrus 0.0509ᵗ
All AWM > All Music Majors Right CT Heschl’s Gyrus 0.0001**
All Architecture Majors > All AWM Right WM Heschl’s Gyrus 0.0275*
Right CT Parahippocampal
Gyrus
0.0312*
Right GM Hippocampus 0.0288*
Right FA Hippocampus 0.0156*
All AWM > All Architecture Majors Left TV POP 0.0922ᵗ
Left GM POP 0.0830ᵗ
Right CT Heschl’s Gyrus 0.0037**
All data analyzed in BrainSuite and statistics processed via SPSS. Table shows only very statistically
significant (**), statistically significant (*), or statistical trend (ᵗ) results. All p-values are corrected for
multiple comparisons using the Holm-Bonferroni adjustment.
AWM compared to architecture majors to p=0.0400*. For the same group comparison,
removal of three architecture majors outliers reduced the significance of the TV result in
left POP to a weak statistical trend (p=0.0920ᵗ), and reduced the significance of the GM
result in left POP to what is arguably no longer a statistical trend (p=0.1000). Finally, in
architecture majors compared to AWM, removal of one architecture major outlier
reduced the significance of the GM result in right hippocampus to just below statistical
significance (p=0.0510ᵗ). All other results not listed above either stayed the same or
improved in statistical significance with the removal of outliers.
124
Table 4: Comparisons Driven by Outliers, Exploratory Analysis 3
Comparison Number
of
Outliers
Group p-value
With
Outliers
p-value
Without
Outliers
All Music Majors > All AWM
TV Right Pars Orbitalis 2 Music Majors 0.0863ᵗ 0.2130
GM Right Pars Orbitalis 1 Music Majors 0.0915ᵗ 0.1600
GM Left Angular Gyrus 1 Music Majors 0.0509ᵗ 0.0850ᵗ
All AWM > All Music Majors
CT Right Heschl’s Gyrus 2 AWM
2 Music Majors 0.0001** 0.0050**
All Architecture Majors > All
AWM
GM Right Hippocampus 1 Architecture
Majors
0.0288* 0.0510ᵗ
All AWM > All Architecture
Majors
CT Right Heschl’s Gyrus 2 AWM
1 Architecture
Majors
0.0037** 0.0400*
TV Left POP 3 Architecture
Majors
0.0922ᵗ 0.0920ᵗ
GM Left POP 3 Architecture
Majors
0.0830ᵗ 0.1000
All data analyzed in BrainSuite and statistics processed via SPSS. Table shows only
results that were negatively affected by the removal of outliers. All p-values are corrected
for multiple comparisons using the Holm-Bonferroni adjustment.
125
ii. Discussion
Given the number of results that decreased in statistical significance when outliers
were removed, the fact that these groups are uneven, and that the AWM group is small
(n=8), reduces the possible meaning of these exploratory comparisons. Notably, the
outliers in question were almost always in either of the two larger groups, music majors
(n=25), or architecture majors (n=16). Removal of outliers in the AWM group did not
reduce any statistically significant results to be below the statistical significance
threshold. Therefore the only result that will not be discussed is that of right pars orbitalis
in music majors compared to AWM subjects.
By examining the AWM group separately I was attempting to conclude if it
represented a true middle group between the music majors and architecture majors. If so,
one would expect results in AWM compared to architecture majors that would reflect, to
a certain degree, the results obtained in music majors compared to architecture majors.
This was the case: the AWM group exhibits increased TV and GM in left POP when
compared to architecture majors. One would not expect to see differences between
architecture majors and AWM subjects; the latter have been subjected to the same
architectural training as the former. Thus, the parahippocampal gyrus and hippocampus
results in this comparison are surprising. Interestingly, the right hippocampus FA result
seen in this comparison is roughly the same as the one present in architecture majors
compared to music majors, and architecture majors compared to vocalists. While no
AWM subjects list voice as their primary instrument, 4 of the 8 list it as their secondary
instrument. Could it be that something about the training and use of the vocal cords alters
white matter microstructure in right hippocampus? Further investigations will be
necessary to address this question.
126
The most intriguing results here are in right Heschl’s gyrus. Heschl’s gyrus,
primary auditory cortex, has repeatedly been said to show larger grey matter volume
and/or density in musicians than non-musicians (reviewed in chapter 1, pg. 30-1). Not
only was this not found in the present study, but in this last comparison, the AWM group
exhibited larger CT in right Heschl’s gyrus than both the architecture majors and the
music majors. On the other hand, the architecture majors exhibited increased WM
volume in right Heschl’s gyrus when compared to the AWM group.
Again, given the reasons listed above and the wide range of musical training
present in the AWM group, it is likely that, while some of these results might reflect the
presence of a true middle group, most of them may simply reflect the small sample size.
Further investigation, controlling for instrumentation (as best as possible), will need to be
conducted to clarify these results.
E. CORRELATIONAL ANALYSES
Correlational analyses were performed between three variables yielded by the
performance background questionnaire, and major (music or architecture), or instrument
group (either instrumentalists or vocalists, to maintain statistical power, given that
correlational analyses are not very meaningful in groups of less than ten subjects) The
first variable to be considered was age of onset of musical training, the age the subject
began playing their first instrument. As the first instrument is often not the primary
instrument in adulthood, the second variable examined was age at which training on the
primary instrument started. The third variable addressed intensity of practice on the
primary instrument, the average hours spent per day in practice of this primary
127
instrument. These answers to all the questions came from self-report by the subjects; thus
subject to possible memory inaccuracies.
I expected to find strong negative correlations between age of onset of musical
training – for both the first and the primary instruments – in the music major group, in at
least motor-related regions. A negative correlation would indicate that these regions
increase in volume, CT, or FA the longer a person has been undergoing musical training.
I also expected to find strong positive correlations between intensity of practice on the
primary instrument and volume, CT, or FA in the same motor-related regions. Results are
presented below and in Table 5 with degrees of freedom, correlation coefficient, and
corrected p-value.
i. Age of onset of use of the first instrument
For music majors, negative correlations were found: in left POP WM, r
s
(23) = -
.490, p=0.075ᵗ. Positive correlations were found: in left precentral gyrus TV, r
s
(23) =
.553, p=0.025*, and GM, r
s
(23) = .504, p=0.048*; and in left insula CT, r
s
(23) = .511,
p=0.055ᵗ.
For instrumentalists, no negative correlations were found. Positive correlations
were found: in left precentral gyrus TV, r
s
(13) = .619, p=0.090ᵗ.
For vocalists, negative correlations were found: in left Broca’s area WM, r
s
(8) = -
.817, p=0.020*; and in right parahippocampal gyrus FA, r
s
(8) = -.773, p=0.045*.
Positive correlations were found: in right POP CT, r
s
(8) = .742, p=0.070ᵗ; and in right
insula TV, r
s
(8) = .823, p=0.015*, and GM, r
s
(8) = .760, p=0.044*.
128
ii. Age of onset of use of the primary instrument
For music majors, negative correlations were found: in left insula WM, r
s
(23) = -
.483, p=0.100ᵗ; and in right precentral gyrus FA, r
s
(23) = -.515, p=0.060ᵗ. No positive
correlations were found.
For instrumentalists, negative correlations were found: in left precentral gyrus
GM, r
s
(13) = -.722, p=0.020*. No positive correlations were found.
For vocalists, negative correlations were found: in left PTR TV, r
s
(8) = -.815,
p=0.028*, WM, r
s
(8) = -.849, p=0.020*, GM, r
s
(8) = -.723, p=0.056ᵗ, and FA, r
s
(8) = -
.740, p=0.069ᵗ; in left Heschl’s gyrus CT, r
s
(8) = -.840, p=0.025*; in right cingulate
gyrus FA, r
s
(23) = -.798, p=0.050*; and in left cingulate gyrus FA, r
s
(23) = -.824,
p=0.030*. No positive correlations were found.
iii. Intensity of practice on the primary instrument
For music majors, negative correlations were found: in right Heschl’s gyrus GM,
r
s
(23) = -.488, p=0.090ᵗ. No positive correlations were found.
For instrumentalists, no negative or positive correlations were found.
For vocalists, negative correlations were found: in right Heschl’s gyrus TV, r
s
(8)
= -.843, p=0.020*, and GM, r
s
(8) = -.758, p=0.072ᵗ. No positive correlations were found.
iv. Discussion
Negative correlations found for age of onset of use of primary instrument accord well
with predictions. As expected, all music majors who started their primary instrument
training at an earlier age exhibited larger WM volumes in left insula and increased FA in
right precentral gyrus. Subdividing further this group showed that all instrumentalists
who started their primary instrument training at an earlier age exhibited larger GM
129
Table 5: Correlational Analyses, by Measurement
Measurement Group Region of Interest Correlation Coefficient p-value
TV
Age of Onset, First Instrument Music Majors Left Precentral Gyrus r
s
(23) = .553 p=0.025*
Instrumentalists Left Precentral Gyrus r
s
(13) = .619 p=0.090ᵗ
Vocalists Right Insula r
s
(08) = .823 p=0.015*
Age of Onset, Primary Instrument Vocalists Left PTR r
s
(08) = -.815 p=0.028*
Intensity of Practice Vocalists Right Heschl’s Gyrus r
s
(08) = -.843 p=0.020*
GM
Age of Onset, First Instrument Music Majors Left Precentral Gyrus r
s
(23) = .504 p=0.048*
Vocalists Right Insula r
s
(08) = .760 p=0.044*
Age of Onset, Primary Instrument Instrumentalists Left Precentral Gyrus r
s
(13) = -.722 p=0.020*
Vocalists Left PTR r
s
(08) = -.723 p=0.056ᵗ
Intensity of Practice Music Majors Right Heschl’s Gyrus r
s
(23) = -.488 p=0.090ᵗ
Vocalists Right Heschl’s Gyrus r
s
(08) = -.758 p=0.072ᵗ
WM
Age of Onset, First Instrument Music Majors Left POP r
s
(23) = -.490 p=0.075ᵗ
Vocalists Left Broca’s Area r
s
(08) = -.817 p=0.020*
Age of Onset, Primary Instrument Music Majors Left Insula r
s
(23) = -.483 p=0.100ᵗ
Vocalists Left PTR r
s
(08) = -.849 p=0.020*
CT
Age of Onset, First Instrument Music Majors Left Insula r
s
(23) = .511 p=0.055ᵗ
Vocalists Right POP r
s
(08) = .742 p=0.070ᵗ
Age of Onset, Primary Instrument Vocalists Left Heschl’s Gyrus r
s
(08) = -.840 p=0.025*
FA
Age of Onset, First Instrument Vocalists Right Parahippocampal Gyrus r
s
(08) = -.773 p=0.045*
Age of Onset, Primary Instrument Music Majors Right Precentral Gyrus r
s
(23) = -.515 p=0.060ᵗ
Vocalists Left PTR r
s
(08) = -.740 p=0.069ᵗ
Vocalists Right Cingulate Gyrus r
s
(23) = -.798 p=0.050*
Vocalists Left Cingulate Gyrus r
s
(23) = -.824 p=0.030*
All statistics processed via SPSS. Table shows only very statistically significant (**), statistically significant (*), or statistical trend (ᵗ) results. All p-values are
corrected for multiple comparisons using the Holm-Bonferroni adjustment.
130
volume in left precentral gyrus, and that all vocalists who started their singing at an
earlier age exhibited larger TV, GM volume, WM volume, and increased FA in left PTR,
increased CT in left Heschl’s gyrus, and increased FA in both right and left cingulate
gyrus. While there were no positive correlations in this analysis, there were both positive
and negative correlations found for the age of onset of use of the first (not necessarily
primary) instrument.
Beginning with the expected negative correlations, all music majors who started
their first instrument training at an earlier age exhibited larger WM volume in left POP.
All vocalists who started their first instrument training at an earlier age exhibited larger
WM volume in left Broca’s area; it is thus likely that the vocalists may be driving the
general music major result. On the other hand, the positive correlations found in this
analysis are intriguing. For music majors, the later the subject started his/her first
instrument, the larger both TV and GM volume were in left precentral gyrus, and the
more increased CT was in left insula. For instrumentalists, the later the subject started
his/her first instrument, the larger TV was in left precentral gyrus. For these same ROIs,
negative correlations exist for the age of first use of the primary instrument. A musician’s
primary instrument is the instrument that typically receives more practice time and
training over the course of his/her lifetime. Most musicians make their first forays into
musical training on the piano or a stringed instrument; thus, many non-pianists or non-
strings-playing musicians have a discrepancy between the age of onset of training for
their first musical instrument, and that of their primary instrument. It is likely, given the
bias toward the primary instrument over time (and its accompanying hand usage
requirements), that the first instrument played may have less to do with the size of left
131
precentral gyrus than the primary instrument. This line of reasoning does not apply to the
positive correlations found in the vocalists, in right insula TV and GM volume, and in
right POP CT. The vocalists, in general, had the greatest range of age of onset, for both
first and primary instrument. One vocalist in particular did not start any musical training
until age 10, while most began in the more typical age 4-7 range. Other explanations for
these positive results could involve individual differences within this fairly small group
(n=10); perhaps the later-starting vocalists within the group are performing more, or
spending more time on their musical training; this could contribute to increased insula
TV and GM volume, due to increased need for vocal control and empathy, and possibly
to increased CT in right POP, in these potentially later-starting vocalists. If the later-
starting vocalists are indeed skewing the results, we should see a positive correlation, as
we do in this analysis. It is difficult to interpret these findings without more background
information, or the ability to split the groups further, by early and late onset of musical
training.
It is possible that the distinction between primary and first instruments, in terms
of changes in ROIs, has to do with consistency of practice over time. Following that line
of thought, however, we expected to see positive correlations between primary
instrument intensity of practice and volumetric, CT, and FA measurements. Positive
correlations in any of the ROIs chosen for these correlational analyses would suggest that
more practice time leads to increases in volume, CT, and/or FA. According to these
results, however, the less time a subject – particularly a vocalist – spends practicing, the
larger TV and GM volume in right Heschl’s gyrus will be. Perhaps this has something to
do with practice strategies. One could argue that, as excess time spent physically
132
practicing the voice is often detrimental to the vocal cords, a thoughtful student might
incorporate more ear training, score-study and/or text translation time, and slightly less
physical practice time, into a daily regimen. Ear training and score-study, particularly
with audio recordings, might help increase volume in right Heschl’s gyrus. Thus, we
would see the positive correlations in TV and GM in right Heschl’s gyrus found here.
This is purely supposition, however, as no data were recorded for non-physical practice
time. It is also possible that these positive correlations in right Heschl’s gyrus have
nothing at all to do with intensity of practice, and that intensity of practice and TV and
GM volume have an inverse relationship in this population for unknown reasons.
Overall, many of the results found in these correlational analyses support findings
present in the extant literature. They offer a glimpse into the potential importance of the
distinction between first instrument and primary instrument when age of onset of musical
training is considered as a variable. Additionally, while average hours spent practicing
daily is a reasonable proxy for intensity of practice, it may be more useful to obtain
information regarding daily/weekly musical training regimens, as these likely vary
widely between individuals, and over the course of time for each musician. Professional
musicians, for example, may stick to very different regimens than conservatory students
(even different than themselves during conservatory days), particularly if they are no
longer studying privately. These are elements to consider in future studies.
F. GENERAL DISCUSSION AND CONCLUSION
The exploratory and correlational analyses provided some insight into the nature
of the results in the planned analysis. For example, as discussed in the instrument group
analysis, while the left PTR portion of the left Broca’s area result (TV, GM, and WM) for
133
music majors compared to architecture majors, is clearly being driven by the music
freshmen group, the left POP portion of it appears to be driven by the instrumentalists.
The vocalists’ contribution to the overall result can be seen in the correlational analyses,
where left PTR (TV, GM, and WM) negatively correlated with primary instrument age of
onset.
Perhaps the span in years studied here is not wide enough to show more
significant volumetric, CT, or FA differences between seniors and freshmen studying
music performance. Given the variability in age range for the onset of musical training
with the primary instrument, and the correlations suggesting that an earlier start in the use
of the primary instrument training yields larger results in left PTR (vocalists), left
precentral gyrus (instrumentalists), left Heschl’s gyrus (vocalists again), left insula (all
music majors), and both right and left cingulate gyrus (once more in vocalists), it seems
likely that the small age difference between the starting and ending of conservatory did
not allow these results to be maintained at a significant level. There are clear structural
and correlational differences between instrumentalists and vocalists, largely in motor-
related areas, as expected. However, correlational differences in bilateral cingulate gyrus
and/or right Heschl’s gyrus might be the result of a different style of musical training
between the two groups. The cingulate gyrus results, in FA, could also simply be
indicative of microstructural differences in white matter between the groups, not
necessarily brought on by musical training. Both of these suggestions should be explored
in the future, with larger numbers of subjects and larger spans in age. In addition, there
could be differences driven by gender in this population of subjects; a gender analysis
was not undertaken for the BrainSuite data, due to time constraints and small sample size.
134
It must be noted that the statistical analyses performed on all BrainSuite data were
very conservative; the reason for such approach was driven by the large variance inside
the groups and the small sample sizes. It is possible that using such conservative statistics
obscured results that could have been significant, or at least be considered strong trends,
in the predicted ROIs. An examination of Tables 6-10 (at the end of this chapter), which
show both corrected and uncorrected p-values, indicates that, if I were to have gone with
the uncorrected p-values, additional significant results and statistical trends would have
been present in the planned analyses in right superior frontal gyrus, right and left middle
frontal gyrus, right and left POP, left PTR, left pars orbitalis, right and left Broca’s area,
right and left inferior frontal gyrus, right precentral gyrus, right and left postcentral gyrus,
left cingulate gyrus, right and left supramarginal gyrus, right angular gyrus, right
Heschl’s gyrus, right and left fusiform gyrus, right and left parahippocampal gyrus, right
and left hippocampus, and right and left insula. It seems possible that, with more subjects
in each group, some of these uncorrected results would reach statistical significance or at
least be considered as a statistical trend. One may ask if, in such a case, the removal of
outliers might not have had a much smaller influence on the level of statistical
significance. In the correlational analyses, prior to applying the Holm-Bonferroni
correction for multiple comparisons, there were quite a few additional strong negative
correlations in the analyses addressing age on onset, and a couple of strong positive
correlations when intensity of practice was considered. The use of a less conservative
statistical procedure might thus have changed the results and overall interpretation of the
dataset, in a way that likely would have better fit my hypotheses and predictions. Some of
the uncorrected p-value results appear to be concordant with the VBM analysis results
135
presented in chapter 2; this will be discussed more thoroughly in the following chapter, in
which I discuss the study in light of all methods performed.
136
Table 6: Planned Analysis, by Major and Year
Region of Interest Type Architecture Majors/Music Majors ARCH Seniors/MUS Seniors ARCH Freshmen/MUS Freshmen
Mean
Difference
Std.
Error
P-Value Corr. P-
Value
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-Value Corr. P-
Value
R Superior Frontal
Gyrus
CT 0.1029 0.0513 0.0553 0.2767 0.1361 0.068 0.0637 0.3186 0.0542 0.0796 0.5148 1
R Superior Frontal
Gyrus
FA 0.0041 0.0106 0.704 1 0.0031 0.0153 0.8396 0.8396 0.0055 0.0147 0.7169 1
R Superior Frontal
Gyrus
R_GM 0 0.0007 0.9491 0.9491 0.0003 0.0009 0.7132 1 -0.0004 0.0011 0.7637 1
R Superior Frontal
Gyrus
R_TV 0.0003 0.0012 0.8245 1 0.0007 0.0016 0.68 1 -0.0003 0.002 0.8946 1
R Superior Frontal
Gyrus
R_WM 0.0002 0.0006 0.7003 1 0.0003 0.0007 0.6841 1 0.0001 0.001 0.9302 0.9302
L Superior Frontal
Gyrus
CT 0.0497 0.0426 0.2521 1 0.0604 0.0458 0.205 0.41 0.038 0.081 0.6488 1
L Superior Frontal
Gyrus
FA 0.0097 0.0095 0.3151 0.9453 0.0101 0.0131 0.4466 0.4466 0.0089 0.0145 0.5503 1
L Superior Frontal
Gyrus
R_GM 0.0005 0.0008 0.5106 0.5106 0.0013 0.0009 0.1951 0.5852 -0.0006 0.0014 0.6757 1
L Superior Frontal
Gyrus
R_TV 0.0013 0.0013 0.3341 0.6682 0.0024 0.0016 0.1439 0.5755 -0.0004 0.0024 0.8823 0.8823
L Superior Frontal
Gyrus
R_WM 0.0008 0.0006 0.2137 1 0.0011 0.0007 0.1433 0.7163 0.0002 0.0011 0.8397 1
R Middle Frontal
Gyrus
CT 0.0081 0.0498 0.8722 0.8722 -0.0419 0.0636 0.5171 1 0.0875 0.083 0.325 1
R Middle Frontal
Gyrus
FA 0.0047 0.0105 0.6575 1 0.0003 0.0147 0.9819 0.9819 0.0112 0.0147 0.4589 0.9179
R Middle Frontal
Gyrus
R_GM 0.0004 0.0006 0.4581 1 0.0013 0.0008 0.12 0.36 -0.001 0.0007 0.1899 0.9493
R Middle Frontal
Gyrus
R_TV 0.0011 0.001 0.2504 1 0.0025 0.0013 0.0699 0.2794 -0.001 0.0013 0.4268 1
R Middle Frontal
Gyrus
R_WM 0.0007 0.0004 0.1092 0.546 0.0012 0.0005 0.0405 0.2026 -0.0001 0.0006 0.8891 0.8891
L Middle Frontal
Gyrus
CT -0.0639 0.065 0.3334 1 -0.043 0.075 0.5741 1 -0.0892 0.1165 0.4624 1
137
Region of Interest Type Architecture Majors/Music Majors ARCH Seniors/MUS Seniors ARCH Freshmen/MUS Freshmen
Mean
Difference
Std.
Error
P-Value Corr. P-
Value
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-Value Corr. P-
Value
L Middle Frontal
Gyrus
FA 0.0036 0.0094 0.7046 0.7046 0.0061 0.0134 0.6548 0.6548 -0.0001 0.0133 0.9918 0.9918
L Middle Frontal
Gyrus
R_GM 0.0004 0.0008 0.593 1 0.0011 0.001 0.2738 1 -0.0007 0.0011 0.517 1
L Middle Frontal
Gyrus
R_TV 0.0008 0.0013 0.5442 1 0.0018 0.0016 0.2793 1 -0.0009 0.0019 0.6308 1
L Middle Frontal
Gyrus
R_WM 0.0004 0.0005 0.5027 1 0.0007 0.0007 0.3149 0.9447 -0.0002 0.0009 0.8316 1
R Pars Opercularis CT 0.0082 0.0761 0.9145 0.9145 0.0876 0.1126 0.4451 1 -0.1137 0.0809 0.182 0.9099
R Pars Opercularis FA 0.0133 0.0092 0.1568 0.784 0.0126 0.0111 0.2683 1 0.0143 0.0169 0.4184 1
R Pars Opercularis R_GM 0.0001 0.0003 0.7284 1 0.0003 0.0004 0.5014 1 -0.0002 0.0003 0.4985 1
R Pars Opercularis R_TV 0.0002 0.0004 0.7196 1 0.0004 0.0006 0.5104 1 -0.0003 0.0005 0.5816 1
R Pars Opercularis R_WM 0.0001 0.0001 0.7341 1 0.0001 0.0002 0.5758 0.5758 -0.0001 0.0002 0.7586 0.7586
L Pars Opercularis CT -0.0017 0.066 0.9793 0.9793 -0.0029 0.0929 0.9752 0.9752 0.0007 0.0939 0.9939 0.9939
L Pars Opercularis FA 0.0081 0.0088 0.3621 0.7241 0.0114 0.0113 0.3266 0.6533 0.0033 0.0149 0.8293 1
L Pars Opercularis R_GM -0.0005 0.0002 0.0268 0.0804 -0.0006 0.0003 0.0937 0.2811 -0.0005 0.0003 0.1324 0.6618
L Pars Opercularis R_TV -0.0008 0.0003 0.0122 0.0608 -0.0008 0.0004 0.044 0.1761 -0.0007 0.0004 0.1497 0.599
L Pars Opercularis R_WM -0.0002 0.0001 0.0168 0.067 -0.0003 0.0001 0.0402 0.201 -0.0002 0.0002 0.2542 0.7626
R Pars Triangularis CT 0.0044 0.0725 0.9515 1 -0.0318 0.0873 0.7198 0.7198 0.0697 0.1184 0.566 1
R Pars Triangularis FA 0.006 0.0102 0.5626 1 0.0051 0.0134 0.7045 1 0.0072 0.0169 0.6787 0.6787
R Pars Triangularis R_GM 0 0.0003 0.9844 0.9844 0.0005 0.0003 0.1437 0.7183 -0.0008 0.0006 0.2362 0.9449
R Pars Triangularis R_TV 0.0001 0.0005 0.8508 1 0.0009 0.0006 0.153 0.612 -0.0011 0.0009 0.23 1
R Pars Triangularis R_WM 0.0001 0.0002 0.6426 1 0.0004 0.0003 0.2201 0.6602 -0.0003 0.0003 0.2767 0.8301
L Pars Triangularis CT 0.026 0.0549 0.6386 0.6386 0.0782 0.0707 0.2804 1 -0.0513 0.0848 0.5554 0.5554
L Pars Triangularis FA 0.0171 0.0106 0.1163 0.349 0.0207 0.0134 0.1373 0.6867 0.0117 0.018 0.532 1
L Pars Triangularis R_GM -0.0007 0.0003 0.0341 0.1707 -0.0002 0.0004 0.6328 1 -0.0014 0.0005 0.0093 0.0467*
L Pars Triangularis R_TV -0.0009 0.0005 0.0655 0.2621 -0.0001 0.0006 0.84 0.84 -0.0021 0.0007 0.0099 0.0396*
L Pars Triangularis R_WM -0.0002 0.0002 0.2152 0.4304 0.0001 0.0002 0.8049 1 -0.0007 0.0003 0.0261 0.0783
R Pars Orbitalis CT 0.0597 0.1377 0.6684 0.6684 0.2414 0.1957 0.2365 0.7095 -0.2341 0.1449 0.133 0.6651
R Pars Orbitalis FA 0.0081 0.0167 0.6306 1 0.0066 0.0224 0.7706 0.7706 0.0103 0.0256 0.6971 0.6971
R Pars Orbitalis R_GM -0.0001 0.0001 0.426 1 -0.0002 0.0002 0.1361 0.6803 0.0001 0.0002 0.5373 1
R Pars Orbitalis R_TV -0.0001 0.0002 0.4664 1 -0.0004 0.0002 0.1421 0.5684 0.0002 0.0003 0.4853 1
R Pars Orbitalis R_WM 0 0.0001 0.6296 1 -0.0001 0.0001 0.245 0.4899 0.0001 0.0001 0.4439 1
L Pars Orbitalis CT -0.0094 0.1107 0.9328 0.9328 0.0756 0.1339 0.5781 0.5781 -0.1352 0.2008 0.5233 1
L Pars Orbitalis FA 0.0069 0.0122 0.5713 1 0.0113 0.0145 0.4423 0.8846 0.0004 0.0215 0.9863 0.9863
L Pars Orbitalis R_GM 0.0002 0.0002 0.356 1 0.0002 0.0002 0.4307 1 0.0002 0.0003 0.637 1
138
Region of Interest Type Architecture Majors/Music Majors ARCH Seniors/MUS Seniors ARCH Freshmen/MUS Freshmen
Mean
Difference
Std.
Error
P-Value Corr. P-
Value
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-Value Corr. P-
Value
L Pars Orbitalis R_TV 0.0003 0.0002 0.2725 1 0.0003 0.0003 0.3562 1 0.0002 0.0004 0.5809 1
L Pars Orbitalis R_WM 0.0001 0.0001 0.1337 0.6683 0.0001 0.0001 0.2236 1 0.0001 0.0001 0.4426 1
R Broca’s Area CT 0.0127 0.1117 0.9102 0.9102 0.0558 0.1426 0.6998 0.6998 -0.044 0.1761 0.8066 0.8066
R Broca’s Area R_GM 0.0001 0.0005 0.8371 1 0.0008 0.0006 0.1818 0.5455 -0.001 0.0007 0.1719 0.5158
R Broca’s Area R_TV 0.0002 0.0007 0.723 1 0.0013 0.0009 0.1744 0.6978 -0.0014 0.0009 0.1553 0.6211
R Broca’s Area R_WM 0.0002 0.0003 0.5914 1 0.0005 0.0004 0.22 0.4401 -0.0004 0.0003 0.1887 0.3774
L Broca’s Area CT 0.0242 0.0958 0.8017 0.8017 0.0752 0.1385 0.5925 0.5925 -0.0505 0.1173 0.6737 0.6737
L Broca’s Area R_GM -0.0012 0.0004 0.005 0.0198* -0.0008 0.0005 0.1742 0.6969 -0.0019 0.0006 0.006 0.0181*
L Broca’s Area R_TV -0.0017 0.0006 0.0074 0.0222* -0.001 0.0008 0.2448 0.7344 -0.0028 0.0008 0.0042 0.0169*
L Broca’s Area R_WM -0.0005 0.0002 0.0379 0.0759 -0.0002 0.0003 0.5019 1 -0.0009 0.0003 0.0116 0.0232*
R Inferior Frontal
Gyrus
CT 0.0723 0.2069 0.7288 1 0.2972 0.2817 0.3045 0.3045 -0.2781 0.2951 0.3654 0.3654
R Inferior Frontal
Gyrus
R_GM 0 0.0004 0.9823 0.9823 0.0005 0.0005 0.2966 0.8899 -0.0009 0.0007 0.2333 0.6999
R Inferior Frontal
Gyrus
R_TV 0.0001 0.0006 0.858 1 0.0009 0.0008 0.2602 1 -0.0012 0.0009 0.2175 0.8701
R Inferior Frontal
Gyrus
R_WM 0.0001 0.0003 0.6485 1 0.0004 0.0003 0.3008 0.6016 -0.0003 0.0003 0.2582 0.5164
L Inferior Frontal
Gyrus
CT 0.0148 0.1732 0.9323 0.9323 0.1508 0.2282 0.5156 1 -0.1857 0.268 0.5054 0.5054
L Inferior Frontal
Gyrus
R_GM -0.001 0.0004 0.0102 0.0408* -0.0006 0.0005 0.2775 1 -0.0018 0.0005 0.0059 0.0176*
L Inferior Frontal
Gyrus
R_TV -0.0014 0.0005 0.0109 0.0327* -0.0007 0.0007 0.3669 1 -0.0026 0.0007 0.0027 0.0106*
L Inferior Frontal
Gyrus
R_WM -0.0004 0.0002 0.0667 0.1334 -0.0001 0.0003 0.7239 0.7239 -0.0009 0.0003 0.0068 0.0137*
R Precentral Gyrus CT -0.0144 0.0619 0.8175 0.8175 -0.0439 0.0903 0.6327 0.6327 0.0363 0.0695 0.6097 1
R Precentral Gyrus FA 0.0152 0.0088 0.0944 0.4722 0.0075 0.0107 0.4922 1 0.0266 0.0154 0.117 0.585
R Precentral Gyrus R_GM -0.0006 0.0004 0.163 0.652 -0.0008 0.0006 0.1938 0.969 -0.0002 0.0005 0.6681 0.6681
R Precentral Gyrus R_TV -0.001 0.0007 0.1667 0.5002 -0.0012 0.0011 0.2618 1 -0.0008 0.001 0.4308 1
R Precentral Gyrus R_WM -0.0005 0.0004 0.3087 0.6175 -0.0004 0.0006 0.4944 0.9889 -0.0006 0.0006 0.3382 1
L Precentral Gyrus CT 0.0256 0.0567 0.655 1 0.0311 0.0739 0.6781 1 0.0184 0.0946 0.8493 0.8493
L Precentral Gyrus FA 0.0071 0.0107 0.5157 1 0.0147 0.0123 0.248 1 -0.0043 0.0189 0.825 1
L Precentral Gyrus R_GM 0.0002 0.0004 0.5818 1 -0.0002 0.0004 0.5457 1 0.001 0.0006 0.1358 0.5431
L Precentral Gyrus R_TV 0.0004 0.0006 0.56 1 -0.0004 0.0007 0.5812 1 0.0016 0.0009 0.1092 0.5459
L Precentral Gyrus R_WM 0.0001 0.0003 0.6683 0.6683 -0.0002 0.0004 0.7284 0.7284 0.0006 0.0004 0.183 0.5491
139
Region of Interest Type Architecture Majors/Music Majors ARCH Seniors/MUS Seniors ARCH Freshmen/MUS Freshmen
Mean
Difference
Std.
Error
P-Value Corr. P-
Value
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-Value Corr. P-
Value
R Postcentral Gyrus CT 0.0269 0.0649 0.6824 0.6824 0.0278 0.0961 0.7775 0.7775 0.0289 0.0711 0.6903 1
R Postcentral Gyrus FA 0.0155 0.009 0.0952 0.4758 0.0117 0.0108 0.2911 1 0.0211 0.0166 0.2339 1
R Postcentral Gyrus R_GM 0.0006 0.0006 0.3167 0.9502 0.0007 0.0008 0.4257 0.8513 0.0005 0.0007 0.5294 1
R Postcentral Gyrus R_TV 0.001 0.0009 0.2967 1 0.0012 0.0013 0.3723 1 0.0006 0.0011 0.6077 1
R Postcentral Gyrus R_WM 0.0004 0.0004 0.3342 0.6683 0.0006 0.0006 0.3536 1 0.0001 0.0005 0.8087 0.8087
L Postcentral Gyrus CT 0.0094 0.0521 0.8591 0.8591 0.0027 0.0613 0.9657 0.9657 0.0195 0.1008 0.8523 1
L Postcentral Gyrus FA 0.0073 0.0112 0.5171 1 0.0108 0.0145 0.4636 1 0.0021 0.0181 0.9118 0.9118
L Postcentral Gyrus R_GM 0.0008 0.001 0.4572 1 -0.0005 0.0015 0.725 1 0.0028 0.0013 0.0792 0.3959
L Postcentral Gyrus R_TV 0.0012 0.0017 0.5017 1 -0.0011 0.0024 0.66 1 0.0047 0.0023 0.0967 0.3869
L Postcentral Gyrus R_WM 0.0004 0.0007 0.5924 1 -0.0006 0.001 0.5853 1 0.0019 0.0011 0.1385 0.4154
R Cingulate Gyrus CT -0.0104 0.0707 0.8839 1 -0.0484 0.0601 0.4319 1 0.0583 0.1518 0.7095 0.7095
R Cingulate Gyrus FA 0.0035 0.0086 0.6817 1 0.0021 0.0124 0.8671 0.8671 0.0057 0.0113 0.6215 1
R Cingulate Gyrus R_GM -0.0001 0.0004 0.7295 1 0.0001 0.0005 0.8213 1 -0.0005 0.0005 0.3393 1
R Cingulate Gyrus R_TV -0.0002 0.0006 0.8062 1 0.0002 0.0009 0.8224 1 -0.0008 0.0007 0.303 1
R Cingulate Gyrus R_WM 0 0.0003 0.9392 0.9392 0.0001 0.0005 0.8506 1 -0.0002 0.0004 0.5415 1
L Cingulate Gyrus CT 0.11 0.0517 0.0423 0.2113 0.0746 0.0567 0.2092 1 0.1709 0.095 0.0987 0.4934
L Cingulate Gyrus FA 0.0034 0.0101 0.7377 0.7377 -0.0037 0.0137 0.7871 0.7871 0.0141 0.014 0.3325 1
L Cingulate Gyrus R_GM -0.0003 0.0004 0.4715 1 -0.0003 0.0006 0.5788 1 -0.0003 0.0006 0.6681 1
L Cingulate Gyrus R_TV -0.0005 0.0007 0.5102 1 -0.0006 0.001 0.5318 1 -0.0003 0.0011 0.8042 1
L Cingulate Gyrus R_WM -0.0002 0.0003 0.6213 1 -0.0003 0.0004 0.5081 1 0 0.0006 0.9735 0.9735
R Supramarginal
Gyrus
CT -0.0358 0.0467 0.4489 0.4489 -0.0259 0.0535 0.6338 0.6338 -0.0478 0.0887 0.6004 1
R Supramarginal
Gyrus
FA 0.0114 0.0094 0.2372 0.4744 0.0165 0.0095 0.0967 0.29 0.0037 0.0195 0.8537 0.8537
R Supramarginal
Gyrus
R_GM -0.001 0.0005 0.0343 0.1715 -0.0011 0.0005 0.0335 0.1675 -0.0008 0.0009 0.4153 1
R Supramarginal
Gyrus
R_TV -0.0015 0.0007 0.0498 0.1991 -0.0017 0.0008 0.0383 0.1534 -0.001 0.0014 0.5032 1
R Supramarginal
Gyrus
R_WM -0.0004 0.0003 0.1394 0.4182 -0.0006 0.0003 0.1024 0.2048 -0.0002 0.0006 0.7073 1
L Supramarginal
Gyrus
CT -0.0121 0.0519 0.8175 0.8175 -0.0352 0.0674 0.6069 0.6069 0.0305 0.0764 0.7012 1
L Supramarginal
Gyrus
FA 0.0037 0.0091 0.6909 1 0.0075 0.0109 0.502 1 -0.002 0.0154 0.8997 0.8997
L Supramarginal
Gyrus
R_GM 0.0006 0.0006 0.3631 1 0.0013 0.0008 0.0951 0.4756 -0.0006 0.0011 0.6351 1
140
Region of Interest Type Architecture Majors/Music Majors ARCH Seniors/MUS Seniors ARCH Freshmen/MUS Freshmen
Mean
Difference
Std.
Error
P-Value Corr. P-
Value
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-Value Corr. P-
Value
L Supramarginal
Gyrus
R_TV 0.0009 0.001 0.3651 1 0.002 0.0013 0.1326 0.5306 -0.0008 0.0016 0.6099 1
L Supramarginal
Gyrus
R_WM 0.0003 0.0004 0.4331 1 0.0007 0.0006 0.2509 0.7526 -0.0003 0.0005 0.5759 1
R Angular Gyrus CT -0.0698 0.0536 0.2023 0.4045 -0.0147 0.0719 0.8412 0.8412 -0.1539 0.0771 0.0668 0.334
R Angular Gyrus FA 0.013 0.0112 0.2568 0.2568 0.024 0.0148 0.1241 0.2482 -0.0036 0.0176 0.8431 1
R Angular Gyrus R_GM -0.001 0.0005 0.0493 0.2463 -0.0015 0.0007 0.0406 0.2032 -0.0002 0.0006 0.7797 1
R Angular Gyrus R_TV -0.0014 0.0007 0.0714 0.2857 -0.0023 0.0011 0.0456 0.1822 0 0.0009 0.9997 0.9997
R Angular Gyrus R_WM -0.0004 0.0003 0.1942 0.5825 -0.0008 0.0004 0.0842 0.2526 0.0002 0.0003 0.5815 1
L Angular Gyrus CT -0.0595 0.0756 0.437 0.8741 -0.1354 0.1031 0.2032 0.4064 0.0639 0.1045 0.5537 0.5537
L Angular Gyrus FA 0.0025 0.0092 0.7854 0.7854 0.0136 0.0101 0.1948 0.7793 -0.014 0.0167 0.4167 1
L Angular Gyrus R_GM -0.0008 0.0005 0.1231 0.6156 -0.0008 0.0006 0.1972 0.5917 -0.0006 0.0008 0.4438 1
L Angular Gyrus R_TV -0.0011 0.0007 0.1284 0.5136 -0.0012 0.0009 0.1912 0.9562 -0.0009 0.0012 0.4718 1
L Angular Gyrus R_WM -0.0003 0.0002 0.186 0.5579 -0.0003 0.0003 0.2443 0.2443 -0.0003 0.0005 0.5343 1
R Superior
Temporal Gyrus
CT -0.0816 0.0563 0.1558 0.7792 -0.0856 0.0736 0.257 1 -0.0671 0.0753 0.3896 1
R Superior
Temporal Gyrus
FA 0.0098 0.0102 0.3432 1 0.0118 0.0116 0.3186 0.9557 0.0069 0.0199 0.7345 0.7345
R Superior
Temporal Gyrus
R_GM -0.0001 0.0004 0.7232 1 -0.0001 0.0006 0.8348 0.8348 -0.0002 0.0005 0.723 1
R Superior
Temporal Gyrus
R_TV 0 0.0006 0.9621 0.9621 0.0002 0.0008 0.8 1 -0.0003 0.0007 0.6764 1
R Superior
Temporal Gyrus
R_WM 0.0002 0.0002 0.3977 1 0.0003 0.0003 0.2404 1 -0.0001 0.0002 0.6798 1
L Superior
Temporal Gyrus
CT -0.023 0.0585 0.6966 1 -0.0402 0.0654 0.5453 1 0.0123 0.1031 0.9073 0.9073
L Superior
Temporal Gyrus
FA 0.0087 0.0101 0.3964 1 0.0129 0.0139 0.364 1 0.0024 0.0153 0.8809 1
L Superior
Temporal Gyrus
R_GM -0.0002 0.0004 0.6033 1 -0.0002 0.0006 0.7777 1 -0.0003 0.0008 0.6666 1
L Superior
Temporal Gyrus
R_TV -0.0001 0.0006 0.861 0.861 0.0001 0.0007 0.9129 0.9129 -0.0004 0.0011 0.7188 1
L Superior
Temporal Gyrus
R_WM 0.0001 0.0002 0.6058 1 0.0002 0.0003 0.3538 1 -0.0001 0.0005 0.8573 1
R Heschl's Gyrus CT 0.1102 0.0614 0.0804 0.4021 0.0883 0.0787 0.2744 0.8231 0.1464 0.1015 0.1717 0.8583
R Heschl's Gyrus FA 0.021 0.0146 0.1602 0.641 0.0211 0.0196 0.2949 0.5898 0.0208 0.023 0.3923 1
141
Region of Interest Type Architecture Majors/Music Majors ARCH Seniors/MUS Seniors ARCH Freshmen/MUS Freshmen
Mean
Difference
Std.
Error
P-Value Corr. P-
Value
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-Value Corr. P-
Value
R Heschl's Gyrus R_GM 0 0.0001 0.7723 0.7723 0 0.0001 0.6998 0.6998 -0.0002 0.0002 0.5001 1
R Heschl's Gyrus R_TV 0.0001 0.0002 0.6536 1 0.0002 0.0002 0.198 0.792 -0.0002 0.0003 0.6248 1
R Heschl's Gyrus R_WM 0.0001 0.0001 0.1657 0.497 0.0002 0.0001 0.0703 0.3516 0 0.0001 0.9115 0.9115
L Heschl's Gyrus CT 0.0117 0.1047 0.9115 0.9115 0.0367 0.1303 0.7811 1 -0.0205 0.1856 0.9152 1
L Heschl's Gyrus FA 0.0118 0.0131 0.3759 1 0.0317 0.019 0.1096 0.5479 -0.0182 0.0144 0.2304 1
L Heschl's Gyrus R_GM -0.0001 0.0002 0.4288 1 -0.0002 0.0002 0.2159 0.8638 0.0001 0.0003 0.8543 1
L Heschl's Gyrus R_TV -0.0001 0.0002 0.4971 1 -0.0003 0.0002 0.2705 0.8114 0 0.0004 0.9036 1
L Heschl's Gyrus R_WM 0 0.0001 0.8779 1 0 0.0001 0.8844 0.8844 0 0.0001 0.9516 0.9516
R Fusiform Gyrus CT 0.023 0.0976 0.8155 0.8155 -0.1038 0.1377 0.4603 0.9206 0.2241 0.1186 0.0876 0.438
R Fusiform Gyrus FA 0.005 0.0114 0.6656 1 0.007 0.0131 0.5999 0.5999 0.002 0.021 0.9276 0.9276
R Fusiform Gyrus R_GM -0.0006 0.0007 0.3695 1 -0.0008 0.001 0.4047 1 -0.0002 0.0008 0.7692 1
R Fusiform Gyrus R_TV -0.0013 0.001 0.2054 0.8218 -0.0014 0.0015 0.3711 1 -0.0013 0.0013 0.3442 1
R Fusiform Gyrus R_WM -0.0007 0.0004 0.1141 0.5704 -0.0005 0.0006 0.3888 1 -0.001 0.0007 0.1539 0.6158
L Fusiform Gyrus CT 0.0508 0.0919 0.5847 1 -0.0123 0.1146 0.9164 0.9164 0.147 0.1571 0.3671 1
L Fusiform Gyrus FA -0.0023 0.0117 0.8461 0.8461 -0.0116 0.0169 0.5038 1 0.0117 0.0157 0.4775 0.955
L Fusiform Gyrus R_GM 0.0003 0.0006 0.622 1 -0.0004 0.0008 0.6728 1 0.0013 0.0008 0.14 0.7001
L Fusiform Gyrus R_TV 0.0004 0.0008 0.6626 1 -0.0004 0.0012 0.7171 1 0.0016 0.0011 0.154 0.616
L Fusiform Gyrus R_WM 0.0001 0.0004 0.8445 1 -0.0001 0.0005 0.8615 1 0.0003 0.0006 0.6016 0.6016
R Parahippocampal
Gyrus
CT 0.2884 0.1322 0.0357 0.1783 0.1716 0.1542 0.2781 1 0.4714 0.2423 0.0731 0.3656
R Parahippocampal
Gyrus
FA 0.0131 0.0137 0.349 1 0.0015 0.0206 0.9434 0.9434 0.0306 0.0159 0.0779 0.3117
R Parahippocampal
Gyrus
R_GM 0.0004 0.0003 0.2005 0.8018 0.0004 0.0004 0.388 0.776 0.0004 0.0004 0.3381 0.6762
R Parahippocampal
Gyrus
R_TV 0.0003 0.0004 0.404 0.8081 0.0005 0.0005 0.3226 1 0 0.0005 0.9394 0.9394
R Parahippocampal
Gyrus
R_WM -0.0001 0.0002 0.7384 0.7384 0.0002 0.0002 0.3736 1 -0.0004 0.0002 0.0835 0.2505
L Parahippocampal
Gyrus
CT -0.0511 0.1604 0.7519 0.7519 -0.237 0.2282 0.3113 1 0.24 0.1965 0.2427 0.4853
L Parahippocampal
Gyrus
FA 0.018 0.0141 0.2133 0.6398 0.0071 0.0185 0.7094 0.7094 0.0344 0.0223 0.1563 0.6254
L Parahippocampal
Gyrus
R_GM -0.0004 0.0003 0.2914 0.5828 -0.0003 0.0005 0.4809 1 -0.0004 0.0005 0.4262 0.4262
L Parahippocampal
Gyrus
R_TV -0.0007 0.0005 0.1523 0.6091 -0.0004 0.0006 0.4835 1 -0.0011 0.0008 0.1872 0.5615
142
Region of Interest Type Architecture Majors/Music Majors ARCH Seniors/MUS Seniors ARCH Freshmen/MUS Freshmen
Mean
Difference
Std.
Error
P-Value Corr. P-
Value
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-Value Corr. P-
Value
L Parahippocampal
Gyrus
R_WM -0.0003 0.0002 0.117 0.5851 -0.0001 0.0002 0.6187 1 -0.0007 0.0004 0.1306 0.6531
R Hippocampus FA 0.0248 0.0081 0.004 0.0160* 0.0302 0.0111 0.0124 0.0497* 0.0168 0.0122 0.1915 0.5745
R Hippocampus R_GM 0.0001 0.0001 0.286 0.858 0 0.0002 0.9177 1 0.0003 0.0002 0.0585 0.2339
R Hippocampus R_TV 0.0001 0.0002 0.5748 1 0 0.0002 0.9627 0.9627 0.0002 0.0002 0.2572 0.5143
R Hippocampus R_WM 0 0.0002 0.81 0.81 0 0.0002 0.8971 1 -0.0001 0.0002 0.5846 0.5846
L Hippocampus FA 0.0185 0.0148 0.2229 0.8915 0.0137 0.0211 0.5271 1 0.0257 0.0207 0.2416 0.9664
L Hippocampus R_GM 0.0001 0.0002 0.5629 0.5629 0 0.0003 0.998 0.998 0.0003 0.0002 0.2955 0.5911
L Hippocampus R_TV 0.0002 0.0002 0.3973 1 0.0001 0.0003 0.8276 1 0.0004 0.0004 0.2861 0.8582
L Hippocampus R_WM 0.0001 0.0002 0.5269 1 0.0001 0.0002 0.6976 1 0.0001 0.0003 0.6482 0.6482
R Insula CT 0.1685 0.1223 0.1759 0.8794 0.0859 0.168 0.6143 1 0.3137 0.1483 0.0563 0.2817
R Insula FA 0.0128 0.0115 0.2766 1 0.0022 0.0134 0.8717 0.8717 0.0286 0.0208 0.2102 0.8409
R Insula R_GM -0.0001 0.0002 0.5856 1 -0.0003 0.0002 0.2189 1 0.0002 0.0002 0.4594 1
R Insula R_TV -0.0001 0.0002 0.5655 1 -0.0003 0.0003 0.2909 1 0.0001 0.0004 0.6769 1
R Insula R_WM 0 0.0001 0.691 0.691 0 0.0001 0.7472 1 0 0.0002 0.8119 0.8119
L Insula CT -0.0099 0.1353 0.942 0.942 -0.2296 0.1691 0.1885 0.9424 0.3495 0.1795 0.0826 0.3305
L Insula FA 0.0022 0.0127 0.8646 1 0.0124 0.0177 0.492 1 -0.0132 0.0184 0.4908 0.9815
L Insula R_GM 0 0.0002 0.755 1 -0.0002 0.0002 0.3159 1 0.0002 0.0003 0.4585 1
L Insula R_TV -0.0001 0.0002 0.7353 1 -0.0001 0.0003 0.7862 0.7862 0 0.0003 0.8918 0.8918
L Insula R_WM 0 0.0001 0.7984 1 0.0001 0.0002 0.5443 1 -0.0002 0.0001 0.0643 0.3213
143
Table 7: Planned Analysis, by Year within Major
Region of Interest Type MUS Freshmen/MUS Seniors
ARCH Freshmen/ARCH Seniors
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
R Superior Frontal Gyrus CT 0.0991 0.0508 0.0645 0.3223 0.0172 0.0915 0.8539 1
R Superior Frontal Gyrus FA 0.0001 0.0151 0.9949 0.9949 0.0024 0.0149 0.8736 1
R Superior Frontal Gyrus R_GM 0.0015 0.0009 0.0985 0.394 0.0008 0.0012 0.5175 1
R Superior Frontal Gyrus R_TV 0.0017 0.0014 0.2387 0.716 0.0007 0.0021 0.7365 1
R Superior Frontal Gyrus R_WM 0.0002 0.0007 0.8279 1 -0.0001 0.001 0.9455 0.9455
L Superior Frontal Gyrus CT 0.0929 0.0547 0.1117 0.5584 0.0704 0.0752 0.3751 1
L Superior Frontal Gyrus FA 0.0033 0.0129 0.8025 0.8025 0.0021 0.0146 0.8888 0.8888
L Superior Frontal Gyrus R_GM 0.0016 0.0011 0.1779 0.7117 -0.0003 0.0013 0.8354 1
L Superior Frontal Gyrus R_TV 0.0018 0.0019 0.3414 1 -0.0009 0.0022 0.6976 1
L Superior Frontal Gyrus R_WM 0.0002 0.0009 0.7903 1 -0.0006 0.001 0.568 1
R Middle Frontal Gyrus CT -0.0595 0.0563 0.3013 0.904 0.0699 0.0881 0.4483 0.4483
R Middle Frontal Gyrus FA 0.0034 0.0148 0.821 0.821 0.0143 0.0145 0.3455 0.6909
R Middle Frontal Gyrus R_GM 0.0014 0.0007 0.0703 0.3517 -0.0009 0.0008 0.2781 0.8342
R Middle Frontal Gyrus R_TV 0.0016 0.0012 0.1889 0.7554 -0.0019 0.0014 0.1851 0.7403
R Middle Frontal Gyrus R_WM 0.0002 0.0005 0.7021 1 -0.0011 0.0007 0.136 0.6799
L Middle Frontal Gyrus CT 0.1433 0.0771 0.082 0.4098 0.0971 0.1152 0.4203 0.8406
L Middle Frontal Gyrus FA 0.0019 0.0127 0.8819 0.8819 -0.0043 0.014 0.7631 0.7631
L Middle Frontal Gyrus R_GM -0.0006 0.001 0.5894 1 -0.0024 0.0011 0.0431 0.2156
L Middle Frontal Gyrus R_TV -0.0011 0.0018 0.5296 1 -0.0038 0.0017 0.0441 0.1766
L Middle Frontal Gyrus R_WM -0.0006 0.0008 0.4913 1 -0.0014 0.0007 0.0658 0.1975
R Pars Opercularis CT 0.1553 0.0983 0.128 0.64 -0.046 0.0977 0.6449 1
R Pars Opercularis FA 0.0002 0.0123 0.9871 0.9871 0.0019 0.0161 0.9075 0.9075
R Pars Opercularis R_GM 0.0001 0.0003 0.8158 1 -0.0004 0.0004 0.3239 0.9716
R Pars Opercularis R_TV -0.0002 0.0005 0.7356 1 -0.0008 0.0006 0.2033 0.8131
144
Region of Interest Type MUS Freshmen/MUS Seniors
ARCH Freshmen/ARCH Seniors
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
R Pars Opercularis R_WM -0.0002 0.0002 0.2174 0.8695 -0.0004 0.0002 0.0886 0.4429
L Pars Opercularis CT 0.0068 0.0836 0.9356 1 0.0105 0.1022 0.9198 0.9198
L Pars Opercularis FA 0.0005 0.0121 0.9692 0.9692 -0.0076 0.0143 0.6069 1
L Pars Opercularis R_GM 0.0004 0.0003 0.2312 1 0.0005 0.0003 0.1835 0.9177
L Pars Opercularis R_TV 0.0004 0.0004 0.3633 1 0.0005 0.0004 0.2335 0.9339
L Pars Opercularis R_WM 0 0.0002 0.9073 1 0.0001 0.0002 0.5941 1
R Pars Triangularis CT 0.0721 0.0984 0.475 1 0.1737 0.1094 0.1391 0.4173
R Pars Triangularis FA 0.003 0.015 0.8448 1 0.005 0.0155 0.7533 0.7533
R Pars Triangularis R_GM 0.0007 0.0006 0.2348 1 -0.0006 0.0004 0.2192 0.4383
R Pars Triangularis R_TV 0.0008 0.0008 0.3402 1 -0.0012 0.0007 0.1267 0.5068
R Pars Triangularis R_WM 0 0.0002 0.8623 0.8623 -0.0007 0.0003 0.0665 0.3327
L Pars Triangularis CT 0.1469 0.0859 0.1022 0.5108 0.0174 0.0693 0.8065 0.8065
L Pars Triangularis FA 0.0165 0.0137 0.2427 0.2427 0.0075 0.0178 0.685 1
L Pars Triangularis R_GM 0.0008 0.0005 0.1204 0.4814 -0.0005 0.0004 0.2291 0.6872
L Pars Triangularis R_TV 0.0011 0.0007 0.1359 0.4078 -0.0009 0.0006 0.1589 0.6356
L Pars Triangularis R_WM 0.0003 0.0002 0.2125 0.4251 -0.0005 0.0003 0.1213 0.6067
R Pars Orbitalis CT 0.1369 0.1359 0.3244 1 -0.3386 0.2021 0.1163 0.5817
R Pars Orbitalis FA -0.021 0.0203 0.3116 1 -0.0173 0.0274 0.5418 0.5418
R Pars Orbitalis R_GM -0.0001 0.0002 0.5666 0.5666 0.0003 0.0002 0.1243 0.4972
R Pars Orbitalis R_TV -0.0002 0.0003 0.4794 0.9587 0.0003 0.0002 0.1386 0.4158
R Pars Orbitalis R_WM -0.0001 0.0001 0.3799 1 0.0001 0.0001 0.2922 0.5844
L Pars Orbitalis CT 0.2381 0.1203 0.0598 0.2992 0.0273 0.2092 0.8995 0.8995
L Pars Orbitalis FA 0.021 0.0229 0.3736 1 0.01 0.0122 0.428 1
L Pars Orbitalis R_GM 0.0002 0.0002 0.5016 1 0.0001 0.0003 0.7362 1
L Pars Orbitalis R_TV 0.0002 0.0003 0.611 1 0.0001 0.0004 0.8696 1
L Pars Orbitalis R_WM 0 0.0001 0.9283 0.9283 0 0.0001 0.7163 1
145
Region of Interest Type MUS Freshmen/MUS Seniors
ARCH Freshmen/ARCH Seniors
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
R Broca’s Area CT 0.2273 0.164 0.1824 0.7297 0.1276 0.1564 0.4306 0.4306
R Broca’s Area R_GM 0.0008 0.0007 0.2712 0.8135 -0.001 0.0006 0.1026 0.2052
R Broca’s Area R_TV 0.0006 0.0009 0.5187 1 -0.0021 0.0009 0.0401 0.1202
R Broca’s Area R_WM -0.0002 0.0003 0.5356 0.5356 -0.0011 0.0004 0.0123 0.0490*
L Broca’s Area CT 0.1537 0.1426 0.293 0.293 0.0279 0.1124 0.8074 1
L Broca’s Area R_GM 0.0011 0.0006 0.0703 0.2811 0 0.0005 0.9884 0.9884
L Broca’s Area R_TV 0.0015 0.0008 0.0867 0.2602 -0.0004 0.0008 0.6385 1
L Broca’s Area R_WM 0.0003 0.0003 0.2555 0.511 -0.0004 0.0003 0.2672 1
R Inferior Frontal Gyrus CT 0.3642 0.2601 0.176 0.7039 -0.211 0.3143 0.514 0.514
R Inferior Frontal Gyrus R_GM 0.0007 0.0006 0.3224 0.6449 -0.0007 0.0006 0.2122 0.4244
R Inferior Frontal Gyrus R_TV 0.0004 0.0008 0.6476 0.6476 -0.0017 0.0009 0.0746 0.2238
R Inferior Frontal Gyrus R_WM -0.0003 0.0003 0.2796 0.8387 -0.001 0.0004 0.0161 0.0645
L Inferior Frontal Gyrus CT 0.3918 0.2169 0.0842 0.1683 0.0552 0.2773 0.8461 0.8461
L Inferior Frontal Gyrus R_GM 0.0013 0.0006 0.0336 0.1344 0.0001 0.0005 0.8233 1
L Inferior Frontal Gyrus R_TV 0.0016 0.0007 0.0452 0.1356 -0.0003 0.0007 0.6296 1
L Inferior Frontal Gyrus R_WM 0.0003 0.0003 0.2466 0.2466 -0.0004 0.0003 0.1474 0.5898
R Precentral Gyrus CT 0.0225 0.0718 0.7567 1 0.1028 0.0884 0.2647 1
R Precentral Gyrus FA -0.007 0.0118 0.5581 1 0.0121 0.0146 0.4316 0.8632
R Precentral Gyrus R_GM 0 0.0006 0.9603 0.9603 0.0006 0.0005 0.2895 0.8685
R Precentral Gyrus R_TV -0.0009 0.0011 0.4381 1 -0.0004 0.001 0.6616 0.6616
R Precentral Gyrus R_WM -0.0008 0.0006 0.1884 0.9421 -0.001 0.0006 0.1152 0.5758
L Precentral Gyrus CT 0.0304 0.0809 0.7109 1 0.0177 0.0887 0.8454 0.8454
L Precentral Gyrus FA -0.006 0.0126 0.6401 1 -0.025 0.0187 0.2171 0.6512
L Precentral Gyrus R_GM 0.0005 0.0006 0.4199 1 0.0017 0.0004 0.0005 0.0027**
L Precentral Gyrus R_TV 0.0002 0.0009 0.8466 0.8466 0.0022 0.0007 0.0057 0.0229*
L Precentral Gyrus R_WM -0.0003 0.0005 0.5037 1 0.0004 0.0004 0.2887 0.5773
146
Region of Interest Type MUS Freshmen/MUS Seniors
ARCH Freshmen/ARCH Seniors
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
R Postcentral Gyrus CT 0.0533 0.0604 0.3915 1 0.0544 0.1031 0.6068 1
R Postcentral Gyrus FA -0.0059 0.0127 0.6491 1 0.0035 0.0152 0.824 1
R Postcentral Gyrus R_GM 0.0005 0.0006 0.4291 1 0.0003 0.0009 0.7141 1
R Postcentral Gyrus R_TV 0.0006 0.001 0.5565 1 0 0.0014 0.9848 0.9848
R Postcentral Gyrus R_WM 0.0001 0.0005 0.8274 0.8274 -0.0003 0.0006 0.5729 1
L Postcentral Gyrus CT -0.015 0.0472 0.7558 0.7558 0.0018 0.1081 0.9869 0.9869
L Postcentral Gyrus FA -0.0075 0.0121 0.5413 1 -0.0163 0.0198 0.4326 1
L Postcentral Gyrus R_GM -0.0025 0.0007 0.0038 0.0113* 0.0009 0.0018 0.6423 1
L Postcentral Gyrus R_TV -0.0045 0.0013 0.0019 0.0074** 0.0012 0.0031 0.7035 1
L Postcentral Gyrus R_WM -0.0021 0.0006 0.0014 0.0070** 0.0003 0.0013 0.8031 1
R Cingulate Gyrus CT 0.0806 0.0893 0.3845 1 0.1874 0.1366 0.2151 0.8602
R Cingulate Gyrus FA -0.0071 0.0141 0.6212 1 -0.0035 0.0091 0.7064 0.7064
R Cingulate Gyrus R_GM 0.0004 0.0005 0.4447 1 -0.0003 0.0006 0.6378 1
R Cingulate Gyrus R_TV 0.0001 0.0008 0.9016 0.9016 -0.0009 0.0009 0.3335 1
R Cingulate Gyrus R_WM -0.0003 0.0004 0.5059 1 -0.0006 0.0004 0.2079 1
L Cingulate Gyrus CT 0.0302 0.0667 0.6586 1 0.1265 0.0883 0.1839 0.7356
L Cingulate Gyrus FA 0.0033 0.0139 0.8128 1 0.0212 0.0137 0.1496 0.7481
L Cingulate Gyrus R_GM 0 0.0006 0.9895 0.9895 0.0001 0.0006 0.8994 1
L Cingulate Gyrus R_TV -0.0005 0.0011 0.6879 1 -0.0001 0.001 0.9014 0.9014
L Cingulate Gyrus R_WM -0.0004 0.0006 0.433 1 -0.0002 0.0004 0.6416 1
R Supramarginal Gyrus CT 0.0675 0.0658 0.3211 1 0.0455 0.0799 0.5834 0.5834
R Supramarginal Gyrus FA -0.0006 0.0135 0.9648 1 -0.0134 0.0169 0.4589 1
R Supramarginal Gyrus R_GM 0.0005 0.0005 0.3007 1 0.0009 0.0009 0.3633 1
R Supramarginal Gyrus R_TV 0.0005 0.0008 0.5097 1 0.0013 0.0014 0.41 1
R Supramarginal Gyrus R_WM 0 0.0004 0.9983 0.9983 0.0004 0.0005 0.5293 1
L Supramarginal Gyrus CT 0.0608 0.0591 0.3146 1 0.1264 0.083 0.1599 0.6397
147
Region of Interest Type MUS Freshmen/MUS Seniors
ARCH Freshmen/ARCH Seniors
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
L Supramarginal Gyrus FA -0.0101 0.0098 0.3116 1 -0.0196 0.0162 0.261 0.522
L Supramarginal Gyrus R_GM 0.001 0.001 0.3338 1 -0.0009 0.0009 0.3711 0.3711
L Supramarginal Gyrus R_TV 0.0008 0.0014 0.6024 1 -0.0021 0.0014 0.1739 0.5216
L Supramarginal Gyrus R_WM -0.0003 0.0005 0.6035 0.6035 -0.0012 0.0006 0.0503 0.2513
R Angular Gyrus CT 0.1183 0.0666 0.0944 0.3774 -0.021 0.0817 0.8011 1
R Angular Gyrus FA 0.0133 0.0162 0.4237 0.4237 -0.0143 0.0164 0.3993 1
R Angular Gyrus R_GM -0.0009 0.0008 0.2621 0.5242 0.0004 0.0005 0.3762 1
R Angular Gyrus R_TV -0.0018 0.0012 0.1322 0.3966 0.0004 0.0007 0.5361 1
R Angular Gyrus R_WM -0.0009 0.0004 0.0451 0.2256 0 0.0003 0.989 0.989
L Angular Gyrus CT -0.0474 0.0959 0.6257 1 0.1519 0.1112 0.1967 0.7869
L Angular Gyrus FA 0.0053 0.0126 0.6789 1 -0.0223 0.0148 0.1695 0.8474
L Angular Gyrus R_GM 0.0004 0.0006 0.4694 1 0.0006 0.0008 0.4393 1
L Angular Gyrus R_TV 0.0005 0.0008 0.577 1 0.0007 0.0012 0.5633 1
L Angular Gyrus R_WM 0.0001 0.0003 0.8382 0.8382 0.0001 0.0005 0.8446 0.8446
R Superior Temporal
Gyrus
CT 0.118 0.0746 0.1282 0.641 0.1365 0.0743 0.0897 0.4486
R Superior Temporal
Gyrus
FA 0.0041 0.0187 0.8282 1 -0.0008 0.0133 0.954 0.954
R Superior Temporal
Gyrus
R_GM 0 0.0005 0.9476 0.9476 -0.0001 0.0006 0.876 1
R Superior Temporal
Gyrus
R_TV -0.0001 0.0006 0.8661 1 -0.0006 0.0008 0.4748 1
R Superior Temporal
Gyrus
R_WM -0.0001 0.0002 0.7406 1 -0.0005 0.0003 0.0968 0.387
L Superior Temporal Gyrus CT 0.1002 0.0756 0.1997 0.9987 0.1527 0.0959 0.1527 0.7636
L Superior Temporal Gyrus FA 0.002 0.0128 0.8803 1 -0.0086 0.0162 0.6104 1
L Superior Temporal Gyrus R_GM 0.0003 0.0007 0.7256 1 0.0001 0.0006 0.8987 0.8987
148
Region of Interest Type MUS Freshmen/MUS Seniors
ARCH Freshmen/ARCH Seniors
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
L Superior Temporal Gyrus R_TV 0.0002 0.001 0.8588 1 -0.0003 0.0009 0.7297 1
L Superior Temporal Gyrus R_WM -0.0001 0.0004 0.8806 0.8806 -0.0004 0.0003 0.2347 0.9386
R Heschl's Gyrus CT -0.0041 0.1004 0.9683 0.9683 0.054 0.0801 0.5121 1
R Heschl's Gyrus FA 0.0107 0.0172 0.5394 1 0.0105 0.0249 0.6827 1
R Heschl's Gyrus R_GM 0.0003 0.0002 0.1516 0.7578 0.0001 0.0001 0.4182 1
R Heschl's Gyrus R_TV 0.0004 0.0003 0.1956 0.7824 0 0.0002 0.9979 0.9979
R Heschl's Gyrus R_WM 0.0001 0.0001 0.422 1 -0.0001 0.0001 0.3628 1
L Heschl's Gyrus CT 0.1414 0.1126 0.2221 0.8884 0.0842 0.1969 0.6793 1
L Heschl's Gyrus FA 0.0246 0.0163 0.1448 0.7239 -0.0253 0.0173 0.168 0.8399
L Heschl's Gyrus R_GM -0.0001 0.0002 0.5886 1 0.0002 0.0003 0.5319 1
L Heschl's Gyrus R_TV -0.0001 0.0002 0.7088 1 0.0002 0.0004 0.6037 1
L Heschl's Gyrus R_WM 0 0.0001 0.8888 0.8888 0 0.0001 0.8952 0.8952
R Fusiform Gyrus CT -0.1514 0.1114 0.1876 0.9378 0.1765 0.1436 0.2399 1
R Fusiform Gyrus FA -0.0135 0.0128 0.3052 1 -0.0186 0.0212 0.4031 1
R Fusiform Gyrus R_GM -0.0003 0.0009 0.7576 1 0.0003 0.0009 0.7117 1
R Fusiform Gyrus R_TV -0.0001 0.0014 0.969 0.969 0.0001 0.0014 0.9538 0.9538
R Fusiform Gyrus R_WM 0.0002 0.0006 0.7334 1 -0.0003 0.0007 0.6962 1
L Fusiform Gyrus CT -0.1346 0.119 0.28 0.56 0.0246 0.1538 0.8754 1
L Fusiform Gyrus FA -0.0187 0.011 0.1036 0.4144 0.0046 0.0203 0.8229 1
L Fusiform Gyrus R_GM -0.0012 0.0006 0.0725 0.3624 0.0005 0.001 0.6405 1
L Fusiform Gyrus R_TV -0.0016 0.001 0.1341 0.4023 0.0005 0.0012 0.6925 1
L Fusiform Gyrus R_WM -0.0004 0.0006 0.5358 0.5358 0 0.0005 0.9358 0.9358
R Parahippocampal Gyrus CT -0.1761 0.2478 0.4875 0.9751 0.1237 0.1452 0.4155 1
R Parahippocampal Gyrus FA -0.0247 0.0141 0.0939 0.4693 0.0043 0.0218 0.8459 0.8459
R Parahippocampal Gyrus R_GM -0.0003 0.0003 0.3576 1 -0.0003 0.0005 0.5819 1
R Parahippocampal Gyrus R_TV -0.0001 0.0005 0.8368 0.8368 -0.0007 0.0006 0.2499 0.9996
149
Region of Interest Type MUS Freshmen/MUS Seniors
ARCH Freshmen/ARCH Seniors
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
R Parahippocampal Gyrus R_WM 0.0002 0.0002 0.4777 1 -0.0004 0.0002 0.0266 0.1328
L Parahippocampal Gyrus CT -0.2795 0.2109 0.1987 0.5961 0.1974 0.215 0.374 1
L Parahippocampal Gyrus FA -0.0284 0.0146 0.0692 0.3461 -0.0011 0.025 0.9669 0.9669
L Parahippocampal Gyrus R_GM -0.0001 0.0006 0.8959 0.8959 -0.0002 0.0004 0.6816 1
L Parahippocampal Gyrus R_TV 0.0004 0.0008 0.6019 1 -0.0002 0.0006 0.681 1
L Parahippocampal Gyrus R_WM 0.0005 0.0004 0.185 0.7398 -0.0001 0.0003 0.8321 1
R Hippocampus FA 0.0102 0.0122 0.415 0.83 -0.0033 0.0111 0.7699 1
R Hippocampus R_GM -0.0003 0.0001 0.0461 0.1842 0.0001 0.0002 0.6058 1
R Hippocampus R_TV -0.0001 0.0002 0.7863 0.7863 0.0001 0.0002 0.4888 1
R Hippocampus R_WM 0.0002 0.0002 0.3159 0.9477 0 0.0002 0.8506 0.8506
L Hippocampus FA 0.0009 0.0159 0.9563 0.9563 0.0128 0.0249 0.616 1
L Hippocampus R_GM -0.0005 0.0002 0.0597 0.2386 -0.0002 0.0003 0.3954 1
L Hippocampus R_TV -0.0006 0.0004 0.153 0.4591 -0.0003 0.0003 0.3631 1
L Hippocampus R_WM -0.0001 0.0003 0.6629 1 0 0.0003 0.8836 0.8836
R Insula CT 0.1105 0.1759 0.5362 1 0.3383 0.1388 0.0342 0.1711
R Insula FA -0.0055 0.0139 0.6967 1 0.021 0.0205 0.3418 0.6835
R Insula R_GM -0.0001 0.0003 0.7882 0.7882 0.0004 0.0002 0.0819 0.3276
R Insula R_TV -0.0001 0.0004 0.7465 1 0.0003 0.0002 0.2238 0.6715
R Insula R_WM -0.0001 0.0002 0.7385 1 -0.0001 0.0001 0.6834 0.6834
L Insula CT -0.1006 0.1539 0.5201 1 0.4785 0.1926 0.0305 0.1523
L Insula FA 0.012 0.0156 0.4524 1 -0.0136 0.0202 0.513 1
L Insula R_GM 0 0.0002 0.9442 0.9442 0.0004 0.0002 0.1347 0.5389
L Insula R_TV 0.0003 0.0003 0.3704 1 0.0003 0.0004 0.3955 1
L Insula R_WM 0.0003 0.0001 0.0662 0.331 -0.0001 0.0002 0.5642 0.5642
150
Table 8: Exploratory Analysis 1 (Instrumentation), by Three Groups (Instrumentalists, Vocalists, Architecture Majors)
Region of Interest Type Instrumentalists/Vocalists Instrumentalists/Architecture Majors Vocalists/Architecture Majors
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-Value Corr. P-
Value
R Pars Opercularis CT -0.0752 0.1146 0.5204 1 -0.0556 0.1048 0.6021 1 0.0196 0.1251 0.877 0.877
R Pars Opercularis FA -0.005 0.0125 0.6925 0.6925 -0.0145 0.0108 0.1942 0.9708 -0.0095 0.0122 0.4453 0.8907
R Pars Opercularis R_GM 0.0004 0.0003 0.2463 0.7389 -0.0001 0.0004 0.7821 0.7821 -0.0005 0.0004 0.2542 0.7627
R Pars Opercularis R_TV 0.0006 0.0005 0.1891 0.9455 -0.0002 0.0006 0.7118 1 -0.0008 0.0006 0.1608 0.643
R Pars Opercularis R_WM 0.0002 0.0002 0.1926 0.7704 -0.0001 0.0002 0.595 1 -0.0004 0.0002 0.0736 0.3681
L Pars Opercularis CT -0.026 0.0893 0.7745 0.7745 -0.0047 0.0887 0.958 0.958 0.0213 0.1048 0.8415 0.8415
L Pars Opercularis FA 0.0113 0.0129 0.3914 1 -0.0067 0.0104 0.5286 1 -0.018 0.0129 0.1845 0.3689
L Pars Opercularis R_GM 0.0002 0.0003 0.4189 1 0.0008 0.0003 0.0329 0.0986 0.0006 0.0003 0.0773 0.3093
L Pars Opercularis R_TV 0.0003 0.0004 0.4523 1 0.0011 0.0004 0.0166 0.0663 0.0008 0.0004 0.0505 0.2523
L Pars Opercularis R_WM 0.0001 0.0002 0.6945 1 0.0003 0.0001 0.013 0.0652 0.0002 0.0002 0.1633 0.4898
R Pars Triangularis CT 0.0715 0.0993 0.4827 1 0.0893 0.0866 0.3171 0.9512 0.0178 0.1124 0.8761 1
R Pars Triangularis FA 0.0083 0.0173 0.6394 1 -0.0006 0.011 0.9542 0.9542 -0.0089 0.017 0.6085 1
R Pars Triangularis R_GM -0.0003 0.0006 0.6522 0.6522 -0.0003 0.0004 0.417 0.8341 -0.0001 0.0005 0.9236 0.9236
R Pars Triangularis R_TV -0.0004 0.0008 0.5751 1 -0.0007 0.0006 0.2166 0.8663 -0.0003 0.0008 0.7311 1
R Pars Triangularis R_WM -0.0002 0.0003 0.5118 1 -0.0004 0.0003 0.117 0.5851 -0.0002 0.0003 0.5159 1
L Pars Triangularis CT 0.0041 0.0985 0.9677 0.9677 -0.0178 0.0659 0.7894 0.7894 -0.0219 0.0926 0.8172 1
L Pars Triangularis FA 0.0197 0.0169 0.269 0.807 -0.0063 0.01 0.5406 1 -0.0259 0.0178 0.1678 0.8389
L Pars Triangularis R_GM 0.0005 0.0005 0.3058 0.6117 0.0007 0.0004 0.1094 0.5468 0.0002 0.0005 0.6621 1
L Pars Triangularis R_TV 0.0008 0.0007 0.2567 1 0.0009 0.0006 0.1728 0.6913 0.0001 0.0007 0.9193 0.9193
L Pars Triangularis R_WM 0.0003 0.0003 0.2158 1 0.0002 0.0002 0.4117 1 -0.0001 0.0003 0.643 1
R Pars Orbitalis CT 0.1099 0.1598 0.5039 0.5039 -0.1427 0.1826 0.4497 1 -0.2526 0.2228 0.2723 0.2723
R Pars Orbitalis FA 0.0239 0.0209 0.2654 0.5308 -0.0055 0.0219 0.805 1 -0.0294 0.0216 0.1914 0.3828
R Pars Orbitalis R_GM -0.0004 0.0002 0.0487 0.1461 0 0.0001 0.8347 1 0.0005 0.0002 0.0415 0.1661
R Pars Orbitalis R_TV -0.0007 0.0003 0.0311 0.1245 0 0.0002 0.9887 0.9887 0.0007 0.0003 0.0387 0.1936
R Pars Orbitalis R_WM -0.0002 0.0001 0.0258 0.1292 0 0.0001 0.7455 1 0.0002 0.0001 0.0667 0.2001
L Pars Orbitalis CT -0.0527 0.1532 0.7359 0.7359 -0.0014 0.1177 0.9904 0.9904 0.0513 0.1671 0.7628 0.7628
L Pars Orbitalis FA 0.0284 0.0205 0.1804 0.9018 0.0084 0.0164 0.6125 1 -0.02 0.0167 0.2529 1
L Pars Orbitalis R_GM 0.0001 0.0002 0.6045 1 -0.0001 0.0002 0.7303 1 -0.0002 0.0003 0.4646 0.9293
L Pars Orbitalis R_TV 0.0002 0.0003 0.5002 1 -0.0002 0.0003 0.6097 1 -0.0004 0.0004 0.3228 0.9685
L Pars Orbitalis R_WM 0.0001 0.0001 0.287 1 -0.0001 0.0001 0.3514 1 -0.0002 0.0001 0.1011 0.5054
R Broca’s Area CT -0.0037 0.169 0.9826 0.9826 0.0337 0.1472 0.8211 0.8211 0.0374 0.1681 0.8265 0.8265
151
Region of Interest Type Instrumentalists/Vocalists Instrumentalists/Architecture Majors Vocalists/Architecture Majors
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-Value Corr. P-
Value
R Broca’s Area R_GM 0.0001 0.0007 0.8763 1 -0.0004 0.0006 0.4814 0.9628 -0.0005 0.0007 0.4751 0.9503
R Broca’s Area R_TV 0.0002 0.0009 0.8665 1 -0.001 0.0009 0.2871 0.8614 -0.0011 0.0011 0.3247 0.9742
R Broca’s Area R_WM 0.0001 0.0004 0.8792 1 -0.0005 0.0003 0.1399 0.5598 -0.0006 0.0004 0.2064 0.8255
L Broca’s Area CT -0.022 0.167 0.8974 0.8974 -0.0225 0.1212 0.8544 0.8544 -0.0006 0.1787 0.9974 0.9974
L Broca’s Area R_GM 0.0007 0.0005 0.1923 0.577 0.0015 0.0006 0.0241 0.0964 0.0008 0.0006 0.1736 0.6945
L Broca’s Area R_TV 0.0011 0.0008 0.1715 0.686 0.002 0.0009 0.032 0.0959 0.0009 0.0009 0.316 0.9479
L Broca’s Area R_WM 0.0004 0.0003 0.2544 0.5089 0.0005 0.0003 0.0855 0.1709 0.0001 0.0004 0.7709 1
R Inferior Frontal Gyrus CT 0.1061 0.2946 0.7236 0.7236 -0.109 0.2665 0.6876 0.6876 -0.2151 0.335 0.529 1
R Inferior Frontal Gyrus R_GM -0.0003 0.0006 0.6242 1 -0.0004 0.0005 0.4674 0.9347 -0.0001 0.0006 0.8995 0.8995
R Inferior Frontal Gyrus R_TV -0.0005 0.0008 0.5551 1 -0.0009 0.0008 0.2362 0.7086 -0.0005 0.0009 0.6236 1
R Inferior Frontal Gyrus R_WM -0.0002 0.0003 0.5637 1 -0.0006 0.0003 0.1049 0.4197 -0.0004 0.0004 0.3422 1
L Inferior Frontal Gyrus CT -0.0747 0.2627 0.7801 0.7801 -0.024 0.2074 0.9091 0.9091 0.0507 0.2809 0.8589 1
L Inferior Frontal Gyrus R_GM 0.0008 0.0005 0.129 0.387 0.0014 0.0006 0.0221 0.0882 0.0006 0.0005 0.2834 1
L Inferior Frontal Gyrus R_TV 0.0013 0.0007 0.0895 0.3581 0.0018 0.0008 0.0255 0.0766 0.0005 0.0008 0.512 1
L Inferior Frontal Gyrus R_WM 0.0005 0.0003 0.1563 0.3126 0.0004 0.0002 0.0912 0.1824 -0.0001 0.0004 0.8747 0.8747
R Precentral Gyrus CT 0.0773 0.0746 0.3138 1 0.0839 0.087 0.3498 1 0.0066 0.0953 0.9459 0.9459
R Precentral Gyrus FA 0.0132 0.0119 0.2781 1 -0.005 0.0104 0.6336 1 -0.0182 0.0114 0.128 0.6401
R Precentral Gyrus R_GM 0.0005 0.0006 0.3794 1 0.001 0.0006 0.0975 0.4875 0.0005 0.0005 0.393 1
R Precentral Gyrus R_TV 0.0006 0.0011 0.5845 1 0.0011 0.0011 0.3107 1 0.0005 0.0009 0.5653 1
R Precentral Gyrus R_WM 0.0001 0.0007 0.9298 0.9298 0.0001 0.0006 0.848 0.848 0.0001 0.0006 0.9299 1
L Precentral Gyrus CT -0.0573 0.08 0.4826 1 -0.0418 0.0738 0.5768 1 0.0155 0.083 0.8543 0.8543
L Precentral Gyrus FA 0.0082 0.0119 0.4999 1 -0.0138 0.0128 0.2927 1 -0.022 0.0125 0.0968 0.484
L Precentral Gyrus R_GM -0.0001 0.0006 0.8145 0.8145 0.0004 0.0004 0.4077 1 0.0005 0.0005 0.3152 1
L Precentral Gyrus R_TV -0.0005 0.0008 0.5492 1 0.0003 0.0008 0.7451 1 0.0008 0.0008 0.3732 1
L Precentral Gyrus R_WM -0.0004 0.0005 0.4606 1 -0.0001 0.0004 0.7727 0.7727 0.0003 0.0005 0.6369 1
R Postcentral Gyrus CT -0.0491 0.0534 0.3671 1 -0.0262 0.0993 0.7964 0.7964 0.023 0.0974 0.8177 0.8177
R Postcentral Gyrus FA 0.0149 0.0119 0.2243 1 -0.0081 0.0111 0.4716 1 -0.023 0.0103 0.0391 0.1953
R Postcentral Gyrus R_GM -0.0002 0.0006 0.7876 1 -0.0005 0.0009 0.5547 1 -0.0004 0.0008 0.6654 1
R Postcentral Gyrus R_TV -0.0002 0.001 0.8744 1 -0.001 0.0014 0.4589 1 -0.0009 0.0014 0.5254 1
R Postcentral Gyrus R_WM 0 0.0005 0.9964 0.9964 -0.0005 0.0006 0.3832 1 -0.0005 0.0006 0.4035 1
L Postcentral Gyrus CT -0.0336 0.0416 0.4275 0.4275 -0.0221 0.0639 0.7346 1 0.0115 0.0641 0.8604 0.8604
L Postcentral Gyrus FA 0.0181 0.0132 0.1876 0.9382 -0.0066 0.0132 0.6247 1 -0.0247 0.0149 0.1163 0.5817
L Postcentral Gyrus R_GM 0.0011 0.0009 0.2327 0.9309 0 0.0015 0.9847 1 -0.0011 0.0014 0.436 1
L Postcentral Gyrus R_TV 0.0018 0.0015 0.2419 0.7258 0 0.0025 0.998 0.998 -0.0018 0.0023 0.4342 1
152
Region of Interest Type Instrumentalists/Vocalists Instrumentalists/Architecture Majors Vocalists/Architecture Majors
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-Value Corr. P-
Value
L Postcentral Gyrus R_WM 0.0007 0.0007 0.2907 0.5815 0 0.001 0.9824 1 -0.0007 0.0009 0.4613 0.9227
R Cingulate Gyrus CT -0.0017 0.0721 0.9811 0.9811 0.08 0.0776 0.3134 1 0.0817 0.065 0.2255 1
R Cingulate Gyrus FA 0.0192 0.0152 0.2227 1 0.0028 0.0108 0.8007 1 -0.0165 0.0145 0.2765 1
R Cingulate Gyrus R_GM 0.0004 0.0005 0.4442 1 0.0002 0.0005 0.7295 1 -0.0002 0.0005 0.6954 0.6954
R Cingulate Gyrus R_TV 0.0006 0.0007 0.4224 1 0.0001 0.0009 0.9497 0.9497 -0.0005 0.0008 0.55 1
R Cingulate Gyrus R_WM 0.0002 0.0003 0.5642 1 -0.0001 0.0005 0.8159 1 -0.0003 0.0005 0.4973 1
L Cingulate Gyrus CT 0.0031 0.0527 0.9543 0.9543 -0.0613 0.0672 0.3717 1 -0.0644 0.057 0.2779 1
L Cingulate Gyrus FA 0.0233 0.0138 0.1051 0.5255 0.0144 0.0128 0.2731 1 -0.0089 0.0136 0.5205 1
L Cingulate Gyrus R_GM 0.0001 0.0006 0.9283 1 0.0004 0.0006 0.5715 1 0.0003 0.0006 0.6001 1
L Cingulate Gyrus R_TV 0.0003 0.001 0.7714 1 0.0005 0.0011 0.6071 1 0.0003 0.0009 0.7897 1
L Cingulate Gyrus R_WM 0.0002 0.0005 0.6142 1 0.0002 0.0005 0.7028 0.7028 -0.0001 0.0004 0.8899 0.8899
R Superior Temporal
Gyrus
CT 0.0328 0.0806 0.6886 0.6886 0.1459 0.0732 0.0587 0.2936 0.1131 0.0803 0.1763 0.7052
R Superior Temporal
Gyrus
FA 0.014 0.0175 0.4308 1 -0.0045 0.0126 0.7228 1 -0.0186 0.0151 0.241 0.7231
R Superior Temporal
Gyrus
R_GM 0.0004 0.0005 0.4737 0.9475 0.0003 0.0006 0.6605 1 -0.0001 0.0007 0.8632 0.8632
R Superior Temporal
Gyrus
R_TV 0.0007 0.0006 0.2993 1 0 0.0008 0.9924 0.9924 -0.0006 0.0009 0.4807 0.9614
R Superior Temporal
Gyrus
R_WM 0.0003 0.0002 0.2251 1 -0.0003 0.0003 0.3634 1 -0.0005 0.0003 0.1 0.5
L Superior Temporal
Gyrus
CT 0.0083 0.0757 0.9134 0.9134 0.0836 0.069 0.2383 1 0.0753 0.0676 0.2811 0.8433
L Superior Temporal
Gyrus
FA 0.0199 0.0144 0.1838 0.9191 -0.0042 0.012 0.731 0.731 -0.0241 0.0146 0.1193 0.5966
L Superior Temporal
Gyrus
R_GM 0.0008 0.0007 0.2655 0.7964 0.0006 0.0006 0.3272 1 -0.0002 0.0006 0.7587 0.7587
L Superior Temporal
Gyrus
R_TV 0.0011 0.0009 0.2285 0.914 0.0004 0.0008 0.5823 1 -0.0007 0.0008 0.4227 0.8455
L Superior Temporal
Gyrus
R_WM 0.0004 0.0004 0.308 0.616 -0.0001 0.0003 0.7013 1 -0.0005 0.0003 0.1216 0.4864
R Heschl's Gyrus CT 0.0024 0.0901 0.9791 0.9791 -0.089 0.0876 0.3201 1 -0.0914 0.0844 0.2931 0.8794
R Heschl's Gyrus FA 0.0293 0.0174 0.1075 0.5373 -0.005 0.0184 0.7872 1 -0.0344 0.0193 0.0937 0.3749
R Heschl's Gyrus R_GM 0.0001 0.0002 0.6646 1 0.0001 0.0002 0.479 1 0 0.0002 0.8767 0.8767
R Heschl's Gyrus R_TV 0.0002 0.0003 0.474 1 0 0.0002 0.9954 0.9954 -0.0002 0.0002 0.407 0.8141
153
Region of Interest Type Instrumentalists/Vocalists Instrumentalists/Architecture Majors Vocalists/Architecture Majors
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-Value Corr. P-
Value
R Heschl's Gyrus R_WM 0.0001 0.0001 0.2775 1 -0.0001 0.0001 0.2894 1 -0.0002 0.0001 0.0451 0.2256
L Heschl's Gyrus CT 0.0471 0.1161 0.6887 0.6887 0.0387 0.1314 0.7717 1 -0.0084 0.1303 0.9491 0.9491
L Heschl's Gyrus FA 0.0246 0.0179 0.1847 0.9237 -0.0121 0.0179 0.508 1 -0.0366 0.0192 0.0728 0.364
L Heschl's Gyrus R_GM -0.0002 0.0002 0.4063 1 0.0001 0.0002 0.4807 1 0.0003 0.0002 0.1577 0.6306
L Heschl's Gyrus R_TV -0.0002 0.0002 0.362 1 0.0001 0.0002 0.5855 1 0.0003 0.0002 0.1682 0.5047
L Heschl's Gyrus R_WM -0.0001 0.0001 0.4226 0.8452 0 0.0001 0.8275 0.8275 0.0001 0.0001 0.4835 0.967
R Fusiform Gyrus CT -0.1912 0.1337 0.1778 0.8889 -0.0333 0.1191 0.7845 0.7845 0.1579 0.1631 0.3462 1
R Fusiform Gyrus FA -0.006 0.0124 0.6331 1 -0.0148 0.0136 0.2919 1 -0.0088 0.015 0.565 0.565
R Fusiform Gyrus R_GM -0.0007 0.0009 0.4701 1 0.0005 0.0009 0.6287 1 0.0012 0.0012 0.3279 1
R Fusiform Gyrus R_TV -0.0007 0.0015 0.629 1 0.0011 0.0014 0.4655 1 0.0018 0.0018 0.3152 1
R Fusiform Gyrus R_WM 0 0.0007 0.9433 0.9433 0.0006 0.0006 0.3206 1 0.0007 0.0007 0.3563 0.7126
L Fusiform Gyrus CT -0.164 0.0909 0.0851 0.4256 -0.1072 0.1309 0.4234 1 0.0568 0.115 0.63 1
L Fusiform Gyrus FA 0.0001 0.0118 0.9907 0.9907 0.0042 0.017 0.8081 1 0.0041 0.0174 0.8186 1
L Fusiform Gyrus R_GM -0.0002 0.0007 0.7939 1 -0.0002 0.0008 0.8149 1 0 0.0009 0.9881 0.9881
L Fusiform Gyrus R_TV 0.0003 0.0011 0.7967 1 -0.0001 0.0012 0.955 0.955 -0.0003 0.0012 0.779 1
L Fusiform Gyrus R_WM 0.0005 0.0005 0.369 1 0.0001 0.0005 0.8093 1 -0.0003 0.0005 0.488 1
R Parahippocampal
Gyrus
CT -0.1863 0.2284 0.4238 1 -0.3166 0.1747 0.0848 0.4239 -0.1302 0.1863 0.4968 0.9937
R Parahippocampal
Gyrus
FA -0.0361 0.0142 0.0233 0.1163 -0.0258 0.0196 0.217 0.6509 0.0103 0.0222 0.6507 0.6507
R Parahippocampal
Gyrus
R_GM -0.0001 0.0004 0.8529 1 -0.0005 0.0004 0.2071 0.8286 -0.0004 0.0005 0.362 1
R Parahippocampal
Gyrus
R_TV 0.0001 0.0005 0.8645 0.8645 -0.0006 0.0005 0.305 0.61 -0.0006 0.0006 0.2921 1
R Parahippocampal
Gyrus
R_WM 0.0002 0.0002 0.4799 1 0 0.0002 0.8432 0.8432 -0.0002 0.0002 0.3273 1
L Parahippocampal
Gyrus
CT -0.1529 0.2359 0.526 1 0.064 0.2155 0.7698 0.7698 0.2169 0.2619 0.4187 1
L Parahippocampal
Gyrus
FA -0.0204 0.0146 0.178 0.8898 -0.0266 0.0191 0.1859 0.7438 -0.0062 0.02 0.763 0.763
L Parahippocampal
Gyrus
R_GM -0.0002 0.0006 0.7298 1 0.0002 0.0004 0.618 1 0.0004 0.0006 0.5035 1
L Parahippocampal
Gyrus
R_TV 0 0.0009 0.9659 0.9659 0.0006 0.0006 0.303 0.909 0.0006 0.0009 0.5119 1
154
Region of Interest Type Instrumentalists/Vocalists Instrumentalists/Architecture Majors Vocalists/Architecture Majors
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-Value Corr. P-
Value
L Parahippocampal
Gyrus
R_WM 0.0002 0.0004 0.497 1 0.0004 0.0002 0.1094 0.5472 0.0002 0.0003 0.6237 1
R Hippocampus FA 0.0228 0.012 0.0728 0.2911 -0.017 0.0103 0.1129 0.4516 -0.0399 0.012 0.0043 0.0171*
R Hippocampus R_GM 0 0.0001 0.8055 0.8055 -0.0001 0.0002 0.5137 1 -0.0001 0.0002 0.7072 0.7072
R Hippocampus R_TV 0.0002 0.0002 0.1935 0.5804 0.0001 0.0002 0.7891 0.7891 -0.0002 0.0002 0.4457 1
R Hippocampus R_WM 0.0003 0.0002 0.196 0.3919 0.0002 0.0002 0.4121 1 -0.0001 0.0002 0.6471 1
L Hippocampus FA 0.0141 0.0151 0.3632 1 -0.0078 0.0213 0.7218 1 -0.0218 0.0223 0.3447 1
L Hippocampus R_GM -0.0003 0.0003 0.2389 0.9557 -0.0003 0.0002 0.177 0.7079 0 0.0003 0.9584 0.9584
L Hippocampus R_TV -0.0002 0.0004 0.5653 1 -0.0004 0.0004 0.2759 0.8277 -0.0002 0.0003 0.6193 1
L Hippocampus R_WM 0.0001 0.0002 0.588 0.588 -0.0001 0.0002 0.755 0.755 -0.0002 0.0002 0.3542 1
R Insula CT -0.0885 0.2063 0.6735 1 -0.0771 0.1404 0.5882 1 0.0114 0.1916 0.9535 0.9535
R Insula FA 0.0079 0.0131 0.552 1 -0.0012 0.0139 0.9312 1 -0.0091 0.0102 0.3857 1
R Insula R_GM -0.0002 0.0003 0.47 1 0.0001 0.0002 0.4093 1 0.0003 0.0003 0.2083 1
R Insula R_TV -0.0002 0.0004 0.536 1 0.0002 0.0003 0.5732 1 0.0004 0.0003 0.2789 1
R Insula R_WM 0 0.0001 0.85 0.85 0 0.0001 0.9611 0.9611 0 0.0001 0.7963 1
L Insula CT -0.0534 0.1657 0.7506 1 0.168 0.1617 0.3108 1 0.2214 0.1738 0.219 1
L Insula FA 0.0115 0.0171 0.5101 1 -0.003 0.0167 0.8578 0.8578 -0.0146 0.0201 0.4785 1
L Insula R_GM 0 0.0002 0.9814 0.9814 0.0002 0.0002 0.2154 1 0.0002 0.0002 0.4024 1
L Insula R_TV 0.0002 0.0003 0.6418 1 0.0003 0.0003 0.3949 1 0.0001 0.0004 0.7756 0.7756
L Insula R_WM 0.0002 0.0001 0.3165 1 0.0001 0.0002 0.7532 1 -0.0001 0.0002 0.5987 1
155
Table 9: Exploratory Analysis 1 (Instrumentation), by Instrument (Piano Majors, Strings Majors, Voice Majors)
Region of Interest Type Piano Majors/Strings Majors Piano Majors/Voice Majors Strings Majors/Voice Majors
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
R Pars Opercularis CT 0.1204 0.1304 0.3742 1 -0.003 0.1343 0.9825 0.9825 -0.1234 0.1294 0.3537 0.7073
R Pars Opercularis FA 0.0148 0.0173 0.4181 0.8362 0.0038 0.0177 0.8333 1 -0.0109 0.0132 0.419 0.419
R Pars Opercularis R_GM -0.0002 0.0005 0.7263 0.7263 0.0002 0.0005 0.6443 1 0.0004 0.0003 0.1804 0.5413
R Pars Opercularis R_TV -0.0007 0.0008 0.4165 1 0.0002 0.0007 0.7723 1 0.0009 0.0005 0.0788 0.3152
R Pars Opercularis R_WM -0.0005 0.0003 0.1171 0.5854 0 0.0003 0.9179 1 0.0004 0.0002 0.041 0.2052
L Pars Opercularis CT 0.1191 0.0964 0.2407 0.963 0.0454 0.1029 0.6658 0.6658 -0.0736 0.0991 0.4675 1
L Pars Opercularis FA 0.0232 0.0134 0.1066 0.533 0.0252 0.0141 0.0967 0.4833 0.002 0.0142 0.8896 0.8896
L Pars Opercularis R_GM 0.0001 0.0004 0.8315 0.8315 0.0003 0.0003 0.4481 1 0.0002 0.0003 0.6187 1
L Pars Opercularis R_TV 0.0002 0.0006 0.7982 1 0.0004 0.0005 0.4911 1 0.0002 0.0004 0.6324 1
L Pars Opercularis R_WM 0.0001 0.0002 0.7492 1 0.0001 0.0002 0.6471 1 0 0.0002 0.807 1
R Pars Triangularis CT 0.0372 0.1017 0.7214 0.7214 0.0938 0.1163 0.4337 1 0.0566 0.1086 0.6092 1
R Pars Triangularis FA 0.0256 0.0139 0.088 0.4401 0.0237 0.0171 0.1912 0.9561 -0.002 0.0192 0.9195 0.9195
R Pars Triangularis R_GM 0.0008 0.0006 0.1696 0.3392 0.0002 0.0006 0.7116 1 -0.0006 0.0006 0.3596 1
R Pars Triangularis R_TV 0.0012 0.0007 0.1411 0.5644 0.0003 0.0009 0.7783 1 -0.0009 0.0008 0.2817 1
R Pars Triangularis R_WM 0.0004 0.0002 0.1584 0.4753 0 0.0003 0.9389 0.9389 -0.0003 0.0003 0.2526 1
L Pars Triangularis CT -0.0499 0.1091 0.6565 0.6565 -0.0259 0.1182 0.8301 0.8301 0.024 0.109 0.8284 0.8284
L Pars Triangularis FA 0.0154 0.0108 0.1754 0.8769 0.0289 0.0171 0.1165 0.5827 0.0135 0.0181 0.4682 0.9364
L Pars Triangularis R_GM -0.0006 0.0006 0.3206 0.9618 0.0001 0.0005 0.8087 1 0.0007 0.0005 0.2014 0.6043
L Pars Triangularis R_TV -0.0009 0.0008 0.3116 1 0.0003 0.0008 0.7235 1 0.0012 0.0008 0.1659 0.6636
L Pars Triangularis R_WM -0.0003 0.0003 0.3279 0.6558 0.0002 0.0003 0.5929 1 0.0004 0.0003 0.1428 0.7139
R Pars Orbitalis CT -0.1527 0.1405 0.3026 1 0.0183 0.1837 0.9222 0.9222 0.1709 0.167 0.3231 0.6462
R Pars Orbitalis FA 0.0157 0.0352 0.6693 0.6693 0.0333 0.0351 0.3743 0.7486 0.0176 0.0207 0.4077 0.4077
R Pars Orbitalis R_GM 0.0002 0.0002 0.3931 1 -0.0003 0.0002 0.1753 0.526 -0.0005 0.0002 0.0365 0.1461
R Pars Orbitalis R_TV 0.0002 0.0003 0.3892 1 -0.0005 0.0003 0.0855 0.342 -0.0007 0.0003 0.0303 0.1516
R Pars Orbitalis R_WM 0.0001 0.0001 0.5694 1 -0.0002 0.0001 0.0385 0.1925 -0.0003 0.0001 0.0421 0.1262
L Pars Orbitalis CT 0.1487 0.1583 0.3798 0.7597 0.0365 0.2003 0.8584 0.8584 -0.1122 0.1499 0.4679 1
L Pars Orbitalis FA -0.0237 0.0255 0.3717 1 0.0142 0.0189 0.4664 0.9329 0.0379 0.0269 0.1813 0.9064
L Pars Orbitalis R_GM 0.0002 0.0002 0.3773 1 0.0003 0.0002 0.2801 0.8402 0 0.0003 0.8967 0.8967
L Pars Orbitalis R_TV 0.0003 0.0003 0.3784 1 0.0004 0.0003 0.2215 0.8861 0.0001 0.0004 0.7836 1
L Pars Orbitalis R_WM 0.0001 0.0001 0.5039 0.5039 0.0001 0.0001 0.1463 0.7313 0.0001 0.0001 0.4999 1
R Broca’s Area CT 0.1576 0.2117 0.4711 1 0.0908 0.2038 0.6641 1 -0.0668 0.196 0.7376 1
R Broca’s Area R_GM 0.0006 0.0008 0.4539 1 0.0005 0.0009 0.5858 1 -0.0001 0.0007 0.852 1
R Broca’s Area R_TV 0.0005 0.001 0.6207 1 0.0005 0.0012 0.695 1 0 0.001 0.9621 0.9621
156
Region of Interest Type Piano Majors/Strings Majors Piano Majors/Voice Majors Strings Majors/Voice Majors
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
R Broca’s Area R_WM -0.0001 0.0003 0.7678 0.7678 0 0.0004 0.9987 0.9987 0.0001 0.0004 0.8083 1
L Broca’s Area CT 0.0692 0.1548 0.6639 0.6639 0.0196 0.1935 0.9209 0.9209 -0.0496 0.178 0.7841 0.7841
L Broca’s Area R_GM -0.0005 0.0007 0.5082 1 0.0004 0.0004 0.367 1 0.0009 0.0007 0.2484 0.4969
L Broca’s Area R_TV -0.0007 0.0009 0.4672 1 0.0007 0.0007 0.3608 1 0.0014 0.001 0.1972 0.7886
L Broca’s Area R_WM -0.0002 0.0003 0.4404 1 0.0003 0.0003 0.4585 0.917 0.0005 0.0004 0.2081 0.6242
R Inferior Frontal Gyrus CT 0.0049 0.3132 0.9878 0.9878 0.1091 0.3436 0.7558 1 0.1042 0.3287 0.7552 0.7552
R Inferior Frontal Gyrus R_GM 0.0008 0.0008 0.3683 1 0.0002 0.0009 0.8589 1 -0.0006 0.0006 0.3404 1
R Inferior Frontal Gyrus R_TV 0.0008 0.001 0.4953 1 0 0.0011 0.9717 0.9717 -0.0008 0.0008 0.3564 1
R Inferior Frontal Gyrus R_WM 0 0.0003 0.9006 1 -0.0002 0.0003 0.567 1 -0.0002 0.0003 0.637 1
L Inferior Frontal Gyrus CT 0.2179 0.2906 0.4754 1 0.056 0.3442 0.8734 0.8734 -0.1619 0.2661 0.5525 0.5525
L Inferior Frontal Gyrus R_GM -0.0003 0.0007 0.7155 0.7155 0.0007 0.0005 0.1885 0.5655 0.0009 0.0007 0.2244 0.4489
L Inferior Frontal Gyrus R_TV -0.0004 0.0009 0.6401 1 0.001 0.0007 0.1595 0.6382 0.0015 0.001 0.1435 0.4305
L Inferior Frontal Gyrus R_WM -0.0002 0.0002 0.5207 1 0.0004 0.0003 0.2717 0.5434 0.0005 0.0003 0.1362 0.5447
R Precentral Gyrus CT 0.1596 0.0811 0.072 0.2882 0.1731 0.0829 0.0568 0.2274 0.0135 0.0824 0.8721 1
R Precentral Gyrus FA 0.0299 0.0122 0.0308 0.1541 0.0311 0.0107 0.0117 0.0585 0.0013 0.014 0.9292 0.9292
R Precentral Gyrus R_GM 0.001 0.0009 0.2948 0.8844 0.0012 0.0008 0.2099 0.6298 0.0001 0.0007 0.8577 1
R Precentral Gyrus R_TV 0.0006 0.0019 0.762 0.762 0.001 0.0017 0.5865 1 0.0004 0.0013 0.7807 1
R Precentral Gyrus R_WM -0.0004 0.001 0.6682 1 -0.0002 0.0009 0.8335 0.8335 0.0002 0.0008 0.7586 1
L Precentral Gyrus CT 0.1309 0.1074 0.2574 0.5148 0.0213 0.113 0.8547 0.8547 -0.1097 0.0814 0.1956 0.5868
L Precentral Gyrus FA 0.0209 0.0168 0.2388 0.7164 0.0207 0.0148 0.1933 0.7733 -0.0002 0.0142 0.9898 0.9898
L Precentral Gyrus R_GM 0.0018 0.0006 0.0143 0.0716 0.001 0.0007 0.168 0.8402 -0.0009 0.0005 0.1278 0.6392
L Precentral Gyrus R_TV 0.002 0.001 0.0646 0.2586 0.0007 0.001 0.4885 1 -0.0013 0.0009 0.1697 0.6788
L Precentral Gyrus R_WM 0.0002 0.0006 0.7619 0.7619 -0.0003 0.0007 0.6915 1 -0.0005 0.0005 0.412 0.824
R Postcentral Gyrus CT 0.0948 0.0897 0.3239 0.3239 0.0078 0.088 0.9323 0.9323 -0.0871 0.0528 0.1179 0.5897
R Postcentral Gyrus FA 0.0368 0.0138 0.0205 0.1026 0.037 0.0102 0.0028 0.0141* 0.0002 0.0145 0.9909 0.9909
R Postcentral Gyrus R_GM 0.0018 0.0008 0.0394 0.1576 0.0009 0.0006 0.1914 0.7656 -0.0009 0.0007 0.2125 0.85
R Postcentral Gyrus R_TV 0.0027 0.0013 0.0544 0.1632 0.0015 0.0011 0.2275 0.6825 -0.0012 0.0011 0.2857 0.8572
R Postcentral Gyrus R_WM 0.0009 0.0006 0.1594 0.3189 0.0006 0.0006 0.3771 0.7541 -0.0004 0.0005 0.4786 0.9572
L Postcentral Gyrus CT 0.0871 0.0512 0.1127 0.4509 0.0187 0.0418 0.6628 0.6628 -0.0684 0.0512 0.2012 1
L Postcentral Gyrus FA 0.0293 0.0135 0.0495 0.2475 0.0357 0.0135 0.0191 0.0957 0.0064 0.015 0.6749 0.6749
L Postcentral Gyrus R_GM 0.0012 0.0016 0.4782 1 0.0018 0.0014 0.2335 0.9339 0.0006 0.0011 0.5675 1
L Postcentral Gyrus R_TV 0.0013 0.0028 0.6628 1 0.0026 0.0025 0.3391 1 0.0013 0.0017 0.4635 1
L Postcentral Gyrus R_WM 0.0001 0.0013 0.9358 0.9358 0.0008 0.0012 0.5175 1 0.0007 0.0007 0.3651 1
R Cingulate Gyrus CT -0.1292 0.1108 0.2645 0.7936 -0.0792 0.0819 0.3601 1 0.05 0.0947 0.6077 0.6077
R Cingulate Gyrus FA 0.0151 0.0165 0.3762 0.7525 0.0283 0.0176 0.1317 0.6583 0.0132 0.0169 0.4467 0.8933
R Cingulate Gyrus R_GM -0.0008 0.0006 0.2041 1 -0.0001 0.0005 0.8062 1 0.0007 0.0006 0.2641 1
157
Region of Interest Type Piano Majors/Strings Majors Piano Majors/Voice Majors Strings Majors/Voice Majors
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
R Cingulate Gyrus R_TV -0.0013 0.0011 0.2532 1 -0.0002 0.0007 0.8041 1 0.0011 0.001 0.276 1
R Cingulate Gyrus R_WM -0.0004 0.0005 0.4297 0.4297 -0.0001 0.0004 0.8648 0.8648 0.0004 0.0005 0.4393 1
L Cingulate Gyrus CT -0.1198 0.0765 0.1448 0.4343 -0.0688 0.044 0.1457 0.5828 0.051 0.0738 0.5045 0.5045
L Cingulate Gyrus FA 0.0124 0.0184 0.5108 0.5108 0.0308 0.0166 0.0919 0.4594 0.0184 0.0165 0.2818 1
L Cingulate Gyrus R_GM -0.0015 0.0008 0.0964 0.3855 -0.0008 0.0007 0.2385 0.477 0.0006 0.0007 0.3912 0.7824
L Cingulate Gyrus R_TV -0.0024 0.0013 0.0943 0.4713 -0.0012 0.0009 0.2382 0.7146 0.0013 0.0013 0.3504 1
L Cingulate Gyrus R_WM -0.0009 0.0006 0.1647 0.3293 -0.0003 0.0004 0.424 0.424 0.0006 0.0006 0.3538 1
R Superior Temporal
Gyrus
CT -0.0019 0.1062 0.9861 0.9861 0.0317 0.098 0.7527 0.7527 0.0335 0.0963 0.7321 1
R Superior Temporal
Gyrus
FA 0.0282 0.0185 0.1543 0.7717 0.031 0.0159 0.0731 0.2924 0.0028 0.0215 0.8995 0.8995
R Superior Temporal
Gyrus
R_GM 0.0003 0.0006 0.591 1 0.0006 0.0007 0.395 0.79 0.0002 0.0006 0.6699 1
R Superior Temporal
Gyrus
R_TV 0.0006 0.0007 0.4079 1 0.001 0.0007 0.1966 0.5899 0.0004 0.0007 0.5553 1
R Superior Temporal
Gyrus
R_WM 0.0003 0.0002 0.2285 0.914 0.0004 0.0002 0.0595 0.2975 0.0002 0.0003 0.5343 1
L Superior Temporal
Gyrus
CT -0.0328 0.1057 0.761 1 -0.0114 0.0849 0.8958 0.8958 0.0215 0.0975 0.829 0.829
L Superior Temporal
Gyrus
FA 0.0152 0.0161 0.3602 1 0.029 0.0162 0.0953 0.4765 0.0138 0.0165 0.4158 0.8317
L Superior Temporal
Gyrus
R_GM 0 0.0008 0.9878 0.9878 0.0008 0.0007 0.2839 1 0.0008 0.0008 0.3735 1
L Superior Temporal
Gyrus
R_TV -0.0002 0.0012 0.8788 1 0.001 0.001 0.3208 0.9624 0.0012 0.0012 0.3163 1
L Superior Temporal
Gyrus
R_WM -0.0002 0.0005 0.7335 1 0.0003 0.0004 0.5248 1 0.0004 0.0005 0.3525 1
R Heschl's Gyrus CT -0.0571 0.1349 0.6797 1 -0.0319 0.1165 0.7907 0.7907 0.0252 0.1104 0.8226 1
R Heschl's Gyrus FA 0.023 0.0227 0.3306 1 0.0431 0.0207 0.0608 0.3039 0.0202 0.0207 0.3444 1
R Heschl's Gyrus R_GM 0.0001 0.0003 0.6455 1 0.0002 0.0002 0.5262 1 0 0.0002 0.879 0.879
R Heschl's Gyrus R_TV 0.0002 0.0004 0.7014 1 0.0003 0.0003 0.4125 1 0.0001 0.0003 0.6843 1
R Heschl's Gyrus R_WM 0 0.0001 0.858 0.858 0.0001 0.0001 0.2576 1 0.0001 0.0001 0.4474 1
L Heschl's Gyrus CT -0.0666 0.1664 0.6958 0.6958 0.0072 0.1414 0.9606 0.9606 0.0737 0.1444 0.6174 0.6174
L Heschl's Gyrus FA 0.0155 0.0221 0.4958 1 0.0339 0.0189 0.0962 0.4808 0.0184 0.0223 0.4222 0.8444
L Heschl's Gyrus R_GM 0.0001 0.0002 0.6354 1 -0.0001 0.0002 0.698 1 -0.0002 0.0002 0.3768 1
L Heschl's Gyrus R_TV 0.0002 0.0003 0.425 1 -0.0001 0.0003 0.7748 1 -0.0003 0.0003 0.2876 1
158
Region of Interest Type Piano Majors/Strings Majors Piano Majors/Voice Majors Strings Majors/Voice Majors
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
L Heschl's Gyrus R_WM 0.0001 0.0001 0.1634 0.817 0 0.0001 0.8983 1 -0.0001 0.0001 0.2244 1
R Fusiform Gyrus CT -0.1173 0.0945 0.237 0.9479 -0.2615 0.1342 0.0752 0.3759 -0.1443 0.1457 0.3379 1
R Fusiform Gyrus FA 0.0002 0.0146 0.9909 0.9909 -0.0059 0.0134 0.665 1 -0.0061 0.0151 0.6922 1
R Fusiform Gyrus R_GM -0.0011 0.0009 0.2176 1 -0.0014 0.001 0.1865 0.746 -0.0003 0.0011 0.8095 1
R Fusiform Gyrus R_TV -0.0014 0.0016 0.3976 1 -0.0016 0.0017 0.3763 1 -0.0002 0.0017 0.9109 1
R Fusiform Gyrus R_WM -0.0003 0.0009 0.7441 1 -0.0002 0.0009 0.8077 0.8077 0.0001 0.0007 0.9252 0.9252
L Fusiform Gyrus CT 0.0925 0.1484 0.5438 1 -0.1085 0.1015 0.3171 1 -0.201 0.1268 0.1428 0.7138
L Fusiform Gyrus FA 0.015 0.0163 0.3755 1 0.0092 0.0153 0.5645 1 -0.0059 0.0136 0.6707 1
L Fusiform Gyrus R_GM -0.0001 0.0011 0.9044 0.9044 -0.0003 0.0011 0.8121 1 -0.0001 0.0007 0.8578 0.8578
L Fusiform Gyrus R_TV -0.0006 0.0018 0.7507 1 -0.0001 0.0019 0.9678 0.9678 0.0005 0.001 0.6093 1
L Fusiform Gyrus R_WM -0.0005 0.0009 0.621 1 0.0002 0.0009 0.8364 1 0.0007 0.0005 0.2348 0.9391
R Parahippocampal
Gyrus
CT -0.5829 0.3291 0.1234 0.2469 -0.5361 0.3506 0.1647 0.6588 0.0468 0.2042 0.8216 0.8216
R Parahippocampal
Gyrus
FA -0.0043 0.0141 0.7625 0.7625 -0.0387 0.016 0.0299 0.1494 -0.0343 0.0159 0.0452 0.226
R Parahippocampal
Gyrus
R_GM -0.0007 0.0003 0.064 0.192 -0.0005 0.0004 0.2791 0.5582 0.0002 0.0004 0.5723 1
R Parahippocampal
Gyrus
R_TV -0.0014 0.0005 0.0335 0.134 -0.0007 0.0006 0.2564 0.7692 0.0006 0.0005 0.2258 0.6773
R Parahippocampal
Gyrus
R_WM -0.0007 0.0003 0.0329 0.1643 -0.0003 0.0003 0.3329 0.3329 0.0004 0.0002 0.094 0.3762
L Parahippocampal
Gyrus
CT 0.3106 0.283 0.3045 1 0.0335 0.3174 0.9178 0.9178 -0.2771 0.2391 0.264 0.7921
L Parahippocampal
Gyrus
FA 0.0122 0.0186 0.5216 1 -0.0131 0.017 0.457 1 -0.0253 0.0175 0.1673 0.8363
L Parahippocampal
Gyrus
R_GM -0.0002 0.0006 0.7538 0.7538 -0.0003 0.0007 0.6508 1 -0.0001 0.0006 0.8356 0.8356
L Parahippocampal
Gyrus
R_TV -0.0008 0.0008 0.3612 1 -0.0004 0.001 0.6724 1 0.0003 0.0009 0.7109 1
L Parahippocampal
Gyrus
R_WM -0.0006 0.0004 0.12 0.5998 -0.0001 0.0004 0.7703 1 0.0005 0.0004 0.262 1
R Hippocampus FA -0.0113 0.0124 0.3873 1 0.0161 0.0101 0.1408 0.5632 0.0273 0.0153 0.0929 0.3717
R Hippocampus R_GM -0.0002 0.0002 0.4383 1 -0.0001 0.0002 0.5453 1 0 0.0001 0.849 0.849
R Hippocampus R_TV -0.0002 0.0003 0.5007 1 0.0001 0.0003 0.6525 0.6525 0.0003 0.0002 0.1385 0.4155
R Hippocampus R_WM 0 0.0002 0.8583 0.8583 0.0003 0.0002 0.2558 0.7673 0.0003 0.0002 0.2389 0.4778
159
Region of Interest Type Piano Majors/Strings Majors Piano Majors/Voice Majors Strings Majors/Voice Majors
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
Difference
Std.
Error
P-
Value
Corr. P-
Value
L Hippocampus FA 0.001 0.0185 0.9565 0.9565 0.0147 0.0161 0.3795 0.759 0.0137 0.0188 0.4789 0.9577
L Hippocampus R_GM 0 0.0003 0.905 1 -0.0004 0.0003 0.2395 0.9579 -0.0003 0.0003 0.3322 1
L Hippocampus R_TV -0.0003 0.0005 0.5413 1 -0.0004 0.0005 0.3758 1 -0.0001 0.0005 0.8611 0.8611
L Hippocampus R_WM -0.0003 0.0003 0.372 1 -0.0001 0.0003 0.8325 0.8325 0.0002 0.0003 0.3789 1
R Insula CT -0.2858 0.2441 0.2745 0.2745 -0.26 0.2738 0.3625 1 0.0258 0.2101 0.9037 0.9037
R Insula FA 0.0318 0.0228 0.1913 0.5739 0.027 0.0191 0.2042 1 -0.0048 0.0154 0.7607 1
R Insula R_GM 0.0004 0.0003 0.1628 0.6513 0 0.0003 0.9303 0.9303 -0.0004 0.0003 0.2434 0.9735
R Insula R_TV 0.0006 0.0004 0.1569 0.7846 0.0001 0.0004 0.7466 1 -0.0005 0.0004 0.2413 1
R Insula R_WM 0.0002 0.0002 0.2408 0.4817 0.0001 0.0002 0.5331 1 -0.0001 0.0002 0.4582 1
L Insula CT -0.263 0.2376 0.3014 0.3014 -0.2112 0.2449 0.412 0.412 0.0518 0.1672 0.7604 1
L Insula FA 0.0345 0.0173 0.0745 0.2981 0.0322 0.0202 0.1341 0.4022 -0.0023 0.0176 0.8993 0.8993
L Insula R_GM 0.0004 0.0002 0.0676 0.3378 0.0003 0.0002 0.3058 0.6115 -0.0002 0.0003 0.5464 1
L Insula R_TV 0.0006 0.0003 0.0763 0.2288 0.0005 0.0003 0.1182 0.4727 -0.0001 0.0004 0.8011 1
L Insula R_WM 0.0002 0.0002 0.2435 0.487 0.0003 0.0002 0.1181 0.5903 0.0001 0.0002 0.7162 1
160
Table 10: Exploratory Analysis 2 (Architecture Majors with Significant Musical Training (AWM), by Three Groups (Architecture
Majors, AWM Subjects, Music Majors)
Region of Interest Type Architecture Majors/AWM Subjects AWM Subjects/Music Majors
Mean
difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
difference
Std.
Error
P-
Value
Corr. P-
Value
R Superior Frontal Gyrus CT 0.1076 0.0682 0.134 0.5361 -0.0046 0.0588 0.9389 0.9389
R Superior Frontal Gyrus FA 0.0076 0.0144 0.6049 0.6049 -0.0036 0.0148 0.8123 1
R Superior Frontal Gyrus R_GM 0.0018 0.0011 0.118 0.5898 -0.0018 0.0011 0.1183 0.5913
R Superior Frontal Gyrus R_TV 0.0025 0.0018 0.1915 0.5744 -0.0022 0.0017 0.219 0.8759
R Superior Frontal Gyrus R_WM 0.0006 0.0008 0.4347 0.8694 -0.0004 0.0007 0.5814 1
L Superior Frontal Gyrus CT 0.1231 0.049 0.0215 0.1076 -0.0734 0.0437 0.1139 0.5693
L Superior Frontal Gyrus FA 0.0151 0.0135 0.286 1 -0.0055 0.0135 0.6933 1
L Superior Frontal Gyrus R_GM 0.0004 0.0012 0.7444 1 0.0001 0.0012 0.908 0.908
L Superior Frontal Gyrus R_TV 0.0006 0.0019 0.7776 1 0.0007 0.0019 0.7074 1
L Superior Frontal Gyrus R_WM 0.0002 0.0008 0.8431 0.8431 0.0006 0.0008 0.479 1
R Middle Frontal Gyrus CT 0.0368 0.0626 0.5652 1 -0.0287 0.0566 0.6208 0.6208
R Middle Frontal Gyrus FA 0.0189 0.0142 0.2078 1 -0.0142 0.0144 0.3416 0.6832
R Middle Frontal Gyrus R_GM -0.0002 0.0006 0.7686 0.7686 0.0006 0.0006 0.2815 0.8446
R Middle Frontal Gyrus R_TV -0.0004 0.001 0.7213 1 0.0015 0.0009 0.1279 0.5117
R Middle Frontal Gyrus R_WM -0.0002 0.0005 0.7067 1 0.0009 0.0005 0.0841 0.4204
L Middle Frontal Gyrus CT -0.0015 0.0769 0.9845 0.9845 -0.0624 0.0674 0.3702 1
L Middle Frontal Gyrus FA 0.014 0.0134 0.3168 1 -0.0104 0.0133 0.4494 1
L Middle Frontal Gyrus R_GM 0.0006 0.0011 0.5793 1 -0.0002 0.0011 0.8374 1
L Middle Frontal Gyrus R_TV 0.0009 0.0018 0.6205 1 -0.0001 0.0017 0.9417 0.9417
L Middle Frontal Gyrus R_WM 0.0003 0.0007 0.706 1 0.0001 0.0007 0.8878 1
R Pars Opercularis CT 0.0456 0.0673 0.5055 1 -0.0373 0.067 0.5819 1
R Pars Opercularis FA 0.0136 0.016 0.4148 1 -0.0003 0.0156 0.9848 0.9848
R Pars Opercularis R_GM 0.0002 0.0003 0.4887 1 -0.0001 0.0003 0.6414 1
161
Region of Interest Type Architecture Majors/AWM Subjects AWM Subjects/Music Majors
Mean
difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
difference
Std.
Error
P-
Value
Corr. P-
Value
R Pars Opercularis R_TV 0.0003 0.0005 0.5628 1 -0.0001 0.0004 0.7439 1
R Pars Opercularis R_WM 0.0001 0.0002 0.7485 0.7485 0 0.0002 0.9455 1
L Pars Opercularis CT -0.0441 0.066 0.5116 1 0.0424 0.0582 0.4742 1
L Pars Opercularis FA 0.0024 0.0147 0.8727 0.8727 0.0057 0.0146 0.702 0.702
L Pars Opercularis R_GM -0.0007 0.0003 0.0166 0.083 0.0002 0.0002 0.5133 1
L Pars Opercularis R_TV -0.001 0.0004 0.023 0.0922 0.0002 0.0004 0.5359 1
L Pars Opercularis R_WM -0.0003 0.0002 0.0752 0.2255 0.0001 0.0002 0.6209 1
R Pars Triangularis CT -0.0347 0.0947 0.7191 0.7191 0.0392 0.0879 0.6635 1
R Pars Triangularis FA 0.0118 0.0202 0.575 1 -0.0058 0.0206 0.7835 0.7835
R Pars Triangularis R_GM -0.0005 0.0004 0.1978 0.989 0.0005 0.0004 0.2277 1
R Pars Triangularis R_TV -0.0009 0.0008 0.2828 1 0.0009 0.0007 0.2302 0.9209
R Pars Triangularis R_WM -0.0003 0.0004 0.4398 1 0.0004 0.0004 0.2842 0.8525
L Pars Triangularis CT 0.0502 0.0776 0.5323 0.5323 -0.0242 0.0837 0.7768 0.7768
L Pars Triangularis FA 0.0234 0.0183 0.229 0.4581 -0.0063 0.0181 0.737 1
L Pars Triangularis R_GM -0.0009 0.0006 0.1542 0.4625 0.0002 0.0006 0.743 1
L Pars Triangularis R_TV -0.0013 0.0009 0.1459 0.7295 0.0004 0.0009 0.6174 1
L Pars Triangularis R_WM -0.0005 0.0003 0.1538 0.6151 0.0002 0.0003 0.4367 1
R Pars Orbitalis CT 0.0783 0.1957 0.6949 0.6949 -0.0186 0.1706 0.9153 0.9153
R Pars Orbitalis FA 0.0418 0.0264 0.1412 0.2824 -0.0337 0.0255 0.2159 0.4319
R Pars Orbitalis R_GM 0.0002 0.0001 0.0862 0.3448 -0.0003 0.0001 0.0183 0.0915
R Pars Orbitalis R_TV 0.0003 0.0002 0.0843 0.4214 -0.0004 0.0002 0.0216 0.0863
R Pars Orbitalis R_WM 0.0001 0.0001 0.1186 0.3559 -0.0001 0.0001 0.0499 0.1496
L Pars Orbitalis CT -0.174 0.1417 0.2376 1 0.1646 0.1296 0.2271 1
L Pars Orbitalis FA -0.0015 0.0258 0.9538 0.9538 0.0085 0.0272 0.7621 1
L Pars Orbitalis R_GM 0.0002 0.0002 0.342 1 0 0.0002 0.829 1
L Pars Orbitalis R_TV 0.0003 0.0003 0.4242 1 0 0.0003 0.9919 0.9919
162
Region of Interest Type Architecture Majors/AWM Subjects AWM Subjects/Music Majors
Mean
difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
difference
Std.
Error
P-
Value
Corr. P-
Value
L Pars Orbitalis R_WM 0.0001 0.0001 0.6743 1 0 0.0001 0.6849 1
R Broca’s Area CT 0.0109 0.1256 0.9322 0.9322 0.0018 0.1275 0.9888 0.9888
R Broca’s Area R_GM -0.0003 0.0006 0.6086 1 0.0004 0.0006 0.4996 0.9993
R Broca’s Area R_TV -0.0006 0.001 0.5972 1 0.0008 0.001 0.4224 1
R Broca’s Area R_WM -0.0002 0.0005 0.606 1 0.0004 0.0004 0.3702 1
L Broca’s Area CT 0.0061 0.1002 0.9521 0.9521 0.0181 0.1066 0.8667 0.8667
L Broca’s Area R_GM -0.0016 0.0007 0.0367 0.1467 0.0003 0.0006 0.6029 1
L Broca’s Area R_TV -0.0024 0.001 0.0435 0.1305 0.0007 0.001 0.5163 1
L Broca’s Area R_WM -0.0008 0.0004 0.0696 0.1391 0.0003 0.0004 0.4178 1
R Inferior Frontal Gyrus CT 0.0891 0.2832 0.7578 1 -0.0168 0.2698 0.9514 0.9514
R Inferior Frontal Gyrus R_GM -0.0001 0.0006 0.8503 0.8503 0.0001 0.0006 0.8639 1
R Inferior Frontal Gyrus R_TV -0.0002 0.001 0.8074 1 0.0004 0.0009 0.712 1
R Inferior Frontal Gyrus R_WM -0.0001 0.0005 0.7686 1 0.0003 0.0004 0.5612 1
L Inferior Frontal Gyrus CT -0.1679 0.2108 0.438 0.438 0.1827 0.2046 0.3865 1
L Inferior Frontal Gyrus R_GM -0.0013 0.0007 0.0892 0.3567 0.0003 0.0007 0.6805 0.6805
L Inferior Frontal Gyrus R_TV -0.0021 0.0011 0.094 0.2819 0.0007 0.0011 0.5578 1
L Inferior Frontal Gyrus R_WM -0.0008 0.0004 0.1191 0.2381 0.0004 0.0004 0.4129 1
R Precentral Gyrus CT -0.0897 0.0838 0.3012 1 0.0753 0.0761 0.343 1
R Precentral Gyrus FA 0.0203 0.0161 0.2378 1 -0.0051 0.0159 0.7535 1
R Precentral Gyrus R_GM -0.0002 0.0006 0.7097 0.7097 -0.0003 0.0006 0.5993 1
R Precentral Gyrus R_TV -0.0006 0.0011 0.6163 1 -0.0005 0.0011 0.6649 1
R Precentral Gyrus R_WM -0.0003 0.0006 0.6345 1 -0.0001 0.0007 0.8248 0.8248
L Precentral Gyrus CT -0.0669 0.1071 0.5466 1 0.0925 0.1059 0.4048 0.8096
L Precentral Gyrus FA 0.0043 0.0174 0.8076 0.8076 0.0027 0.0162 0.87 0.87
L Precentral Gyrus R_GM 0.0007 0.0005 0.1765 0.7059 -0.0005 0.0005 0.3392 1
L Precentral Gyrus R_TV 0.0014 0.0008 0.101 0.5051 -0.0011 0.0008 0.1973 0.9865
163
Region of Interest Type Architecture Majors/AWM Subjects AWM Subjects/Music Majors
Mean
difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
difference
Std.
Error
P-
Value
Corr. P-
Value
L Precentral Gyrus R_WM 0.0008 0.0005 0.1791 0.5372 -0.0006 0.0005 0.269 1
R Postcentral Gyrus CT 0.0373 0.08 0.6463 1 -0.0103 0.0611 0.8686 1
R Postcentral Gyrus FA 0.0202 0.0187 0.3086 1 -0.0047 0.0186 0.807 1
R Postcentral Gyrus R_GM 0.0003 0.0013 0.841 0.841 0.0003 0.0013 0.8165 1
R Postcentral Gyrus R_TV 0.0006 0.002 0.7921 1 0.0004 0.002 0.8366 1
R Postcentral Gyrus R_WM 0.0003 0.0008 0.717 1 0.0001 0.0007 0.8774 0.8774
L Postcentral Gyrus CT 0.0746 0.0829 0.3837 1 -0.0652 0.0711 0.3843 1
L Postcentral Gyrus FA 0.0067 0.0217 0.7635 0.7635 0.0006 0.0207 0.9761 1
L Postcentral Gyrus R_GM 0.0014 0.0011 0.2238 1 -0.0006 0.0008 0.4592 1
L Postcentral Gyrus R_TV 0.0018 0.0019 0.3651 1 -0.0006 0.0014 0.6913 1
L Postcentral Gyrus R_WM 0.0004 0.0009 0.669 1 0 0.0007 0.9861 0.9861
R Cingulate Gyrus CT 0.042 0.0837 0.6221 1 -0.0524 0.0705 0.4705 1
R Cingulate Gyrus FA 0.0166 0.0127 0.2224 1 -0.013 0.0138 0.3628 1
R Cingulate Gyrus R_GM 0.0002 0.0004 0.6133 1 -0.0003 0.0004 0.384 1
R Cingulate Gyrus R_TV 0.0003 0.0007 0.6797 1 -0.0005 0.0007 0.4992 0.9985
R Cingulate Gyrus R_WM 0.0001 0.0004 0.7928 0.7928 -0.0001 0.0004 0.7159 0.7159
L Cingulate Gyrus CT 0.0681 0.0659 0.3163 1 0.0419 0.0576 0.4809 1
L Cingulate Gyrus FA 0.0186 0.0178 0.3224 1 -0.0152 0.0178 0.4148 1
L Cingulate Gyrus R_GM -0.0002 0.0006 0.7093 0.7093 -0.0001 0.0006 0.9165 1
L Cingulate Gyrus R_TV -0.0004 0.001 0.6782 1 -0.0001 0.001 0.9565 1
L Cingulate Gyrus R_WM -0.0002 0.0004 0.6682 1 0 0.0004 0.9753 0.9753
R Supramarginal Gyrus CT 0.0042 0.0828 0.9604 0.9604 -0.04 0.0809 0.6325 1
R Supramarginal Gyrus FA 0.0183 0.0215 0.4175 0.8349 -0.0069 0.0213 0.7529 1
R Supramarginal Gyrus R_GM -0.0009 0.0007 0.2126 0.8505 -0.0001 0.0006 0.8364 0.8364
R Supramarginal Gyrus R_TV -0.0012 0.0009 0.1961 0.9804 -0.0003 0.0008 0.7522 1
R Supramarginal Gyrus R_WM -0.0003 0.0003 0.2627 0.788 -0.0001 0.0002 0.5992 1
164
Region of Interest Type Architecture Majors/AWM Subjects AWM Subjects/Music Majors
Mean
difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
difference
Std.
Error
P-
Value
Corr. P-
Value
L Supramarginal Gyrus CT -0.008 0.0647 0.9028 0.9028 -0.004 0.0596 0.947 0.947
L Supramarginal Gyrus FA 0.0108 0.0142 0.4606 0.9211 -0.0071 0.0131 0.5984 1
L Supramarginal Gyrus R_GM -0.001 0.0013 0.4466 1 0.0016 0.0013 0.2427 0.7282
L Supramarginal Gyrus R_TV -0.0016 0.0019 0.4328 1 0.0025 0.0019 0.2186 0.8743
L Supramarginal Gyrus R_WM -0.0006 0.0007 0.4239 1 0.0009 0.0006 0.1871 0.9355
R Angular Gyrus CT 0.0311 0.0828 0.7138 0.7138 -0.1009 0.0782 0.2253 1
R Angular Gyrus FA 0.0302 0.0202 0.165 0.33 -0.0172 0.0198 0.4057 1
R Angular Gyrus R_GM -0.0013 0.0007 0.0863 0.2588 0.0003 0.0007 0.7243 0.7243
R Angular Gyrus R_TV -0.002 0.0009 0.0548 0.2738 0.0006 0.001 0.5838 1
R Angular Gyrus R_WM -0.0007 0.0004 0.0739 0.2957 0.0003 0.0004 0.4429 1
L Angular Gyrus CT -0.0341 0.0891 0.7066 1 -0.0254 0.0837 0.7658 0.7658
L Angular Gyrus FA 0.019 0.0193 0.3502 1 -0.0165 0.0188 0.406 0.8121
L Angular Gyrus R_GM 0.0004 0.0005 0.4486 1 -0.0011 0.0004 0.0102 0.0509
L Angular Gyrus R_TV 0.0004 0.0007 0.5817 1 -0.0015 0.0006 0.0209 0.0838
L Angular Gyrus R_WM 0 0.0003 0.9521 0.9521 -0.0003 0.0002 0.1791 0.5373
R Temporal Pole CT 0.0811 0.2061 0.7023 1 0.0484 0.2034 0.817 1
R Temporal Pole FA 0.0241 0.0176 0.204 1 -0.0084 0.0179 0.6493 1
R Temporal Pole R_GM -0.0001 0.0009 0.9292 0.9292 -0.0001 0.0009 0.8998 0.8998
R Temporal Pole R_TV -0.0002 0.0011 0.8746 1 -0.0003 0.0011 0.7809 1
R Temporal Pole R_WM -0.0001 0.0003 0.7752 1 -0.0002 0.0004 0.5927 1
L Temporal Pole CT 0.2197 0.1749 0.2362 0.7085 -0.117 0.1705 0.5089 1
L Temporal Pole FA 0.0082 0.0201 0.6902 0.6902 0.0034 0.0192 0.8615 0.8615
L Temporal Pole R_GM 0.0014 0.0006 0.0297 0.1487 -0.0011 0.0004 0.0229 0.1146
L Temporal Pole R_TV 0.0016 0.0009 0.0886 0.3544 -0.0014 0.0007 0.0849 0.3395
L Temporal Pole R_WM 0.0002 0.0004 0.6183 1 -0.0003 0.0004 0.4973 1
R Superior Temporal Gyrus CT -0.1207 0.0801 0.1577 0.7887 0.0391 0.0794 0.6311 1
165
Region of Interest Type Architecture Majors/AWM Subjects AWM Subjects/Music Majors
Mean
difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
difference
Std.
Error
P-
Value
Corr. P-
Value
R Superior Temporal Gyrus FA 0.0192 0.0176 0.3054 1 -0.0094 0.0187 0.626 1
R Superior Temporal Gyrus R_GM -0.0005 0.0005 0.3065 0.9196 0.0004 0.0005 0.4102 1
R Superior Temporal Gyrus R_TV -0.0005 0.0007 0.5176 1 0.0005 0.0006 0.4235 1
R Superior Temporal Gyrus R_WM 0.0001 0.0003 0.7944 0.7944 0.0001 0.0002 0.6758 0.6758
L Superior Temporal Gyrus CT -0.1404 0.0661 0.0481 0.2404 0.1175 0.0621 0.0766 0.3831
L Superior Temporal Gyrus FA 0.0129 0.0156 0.4272 1 -0.0042 0.0155 0.7914 1
L Superior Temporal Gyrus R_GM -0.0002 0.0006 0.6788 1 0 0.0006 0.9819 1
L Superior Temporal Gyrus R_TV -0.0001 0.0007 0.8878 0.8878 0 0.0008 0.998 0.998
L Superior Temporal Gyrus R_WM 0.0001 0.0002 0.528 1 0 0.0002 0.948 1
R Heschl's Gyrus CT -0.2425 0.0599 0.0007 0.0037** 0.3528 0.0632 0 0.0001**
R Heschl's Gyrus FA 0.0224 0.0203 0.2885 0.577 -0.0014 0.019 0.9408 0.9408
R Heschl's Gyrus R_GM 0 0.0001 0.9965 0.9965 0 0.0001 0.816 1
R Heschl's Gyrus R_TV 0.0003 0.0002 0.1604 0.4812 -0.0002 0.0002 0.3512 1
R Heschl's Gyrus R_WM 0.0003 0.0001 0.0069 0.0275* -0.0002 0.0001 0.0714 0.2855
L Heschl's Gyrus CT -0.0024 0.1347 0.9858 0.9858 0.0142 0.1184 0.9067 0.9067
L Heschl's Gyrus FA 0.0152 0.0185 0.4273 1 -0.0034 0.0182 0.8551 1
L Heschl's Gyrus R_GM -0.0004 0.0003 0.1687 0.8437 0.0002 0.0002 0.3323 1
L Heschl's Gyrus R_TV -0.0004 0.0003 0.2096 0.8383 0.0003 0.0003 0.371 1
L Heschl's Gyrus R_WM 0 0.0001 0.6554 1 0 0.0001 0.7245 1
R Fusiform Gyrus CT 0.1691 0.1224 0.1863 0.9314 -0.1461 0.1125 0.2164 0.8655
R Fusiform Gyrus FA 0.0152 0.0158 0.353 1 -0.0102 0.0138 0.4769 0.4769
R Fusiform Gyrus R_GM 0.0005 0.0009 0.5834 1 -0.0011 0.0008 0.2077 1
R Fusiform Gyrus R_TV 0.0002 0.0012 0.8983 0.8983 -0.0015 0.0012 0.2222 0.6666
R Fusiform Gyrus R_WM -0.0003 0.0004 0.4449 1 -0.0004 0.0004 0.3631 0.7262
L Fusiform Gyrus CT 0.0917 0.1039 0.3877 1 -0.0409 0.0882 0.6493 1
L Fusiform Gyrus FA 0.0051 0.0166 0.7627 1 -0.0074 0.0143 0.6161 1
166
Region of Interest Type Architecture Majors/AWM Subjects AWM Subjects/Music Majors
Mean
difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
difference
Std.
Error
P-
Value
Corr. P-
Value
L Fusiform Gyrus R_GM 0.0001 0.001 0.918 0.918 0.0002 0.0009 0.8423 1
L Fusiform Gyrus R_TV 0.0002 0.0014 0.8908 1 0.0002 0.0013 0.896 1
L Fusiform Gyrus R_WM 0.0001 0.0005 0.8614 1 0 0.0005 0.9771 0.9771
R Parahippocampal Gyrus CT 0.6277 0.1789 0.0062 0.0312* -0.3393 0.2012 0.1135 0.5675
R Parahippocampal Gyrus FA 0.0065 0.0172 0.7094 1 0.0066 0.0146 0.6587 1
R Parahippocampal Gyrus R_GM 0.0002 0.0005 0.7023 1 0.0002 0.0005 0.714 0.714
R Parahippocampal Gyrus R_TV 0 0.0008 0.9607 0.9607 0.0004 0.0007 0.6357 1
R Parahippocampal Gyrus R_WM -0.0002 0.0003 0.4274 1 0.0002 0.0003 0.535 1
L Parahippocampal Gyrus CT 0.1542 0.1748 0.3893 1 -0.2053 0.1693 0.2409 0.9638
L Parahippocampal Gyrus FA -0.0008 0.0162 0.9601 0.9601 0.0188 0.0132 0.1748 0.8738
L Parahippocampal Gyrus R_GM -0.0001 0.0004 0.8002 1 -0.0003 0.0004 0.4773 0.9545
L Parahippocampal Gyrus R_TV -0.0003 0.0005 0.5741 1 -0.0005 0.0005 0.3792 1
L Parahippocampal Gyrus R_WM -0.0002 0.0002 0.4607 1 -0.0002 0.0002 0.4878 0.4878
R Hippocampus FA 0.0425 0.0116 0.0039 0.0156* -0.0176 0.012 0.1671 0.5013
R Hippocampus R_GM 0.0003 0.0001 0.0096 0.0288* -0.0002 0.0001 0.0501 0.2002
R Hippocampus R_TV 0.0004 0.0003 0.1223 0.2447 -0.0003 0.0002 0.1964 0.3928
R Hippocampus R_WM 0.0001 0.0002 0.6962 0.6962 -0.0001 0.0002 0.5645 0.5645
L Hippocampus FA 0.0106 0.0178 0.5568 1 0.0079 0.0144 0.592 1
L Hippocampus R_GM 0 0.0003 0.9847 0.9847 0.0001 0.0003 0.7293 1
L Hippocampus R_TV 0.0001 0.0004 0.7155 1 0.0001 0.0004 0.8558 1
L Hippocampus R_WM 0.0001 0.0002 0.5235 1 0 0.0002 0.883 0.883
R Insula CT 0.0151 0.1424 0.9171 0.9171 0.1534 0.1529 0.3299 1
R Insula FA 0.019 0.0225 0.4171 1 -0.0063 0.0218 0.7807 1
R Insula R_GM -0.0003 0.0003 0.3375 1 0.0002 0.0003 0.5282 1
R Insula R_TV -0.0002 0.0003 0.5411 1 0.0001 0.0003 0.8299 0.8299
R Insula R_WM 0.0001 0.0001 0.4878 1 -0.0001 0.0001 0.329 1
167
Region of Interest Type Architecture Majors/AWM Subjects AWM Subjects/Music Majors
Mean
difference
Std.
Error
P-
Value
Corr. P-
Value
Mean
difference
Std.
Error
P-
Value
Corr. P-
Value
L Insula CT -0.1068 0.1678 0.5329 1 0.0969 0.151 0.5321 1
L Insula FA -0.008 0.0222 0.7255 1 0.0102 0.0213 0.6435 1
L Insula R_GM -0.0004 0.0002 0.1344 0.6719 0.0003 0.0002 0.1826 0.9132
L Insula R_TV -0.0004 0.0003 0.3034 1 0.0003 0.0003 0.4077 1
L Insula R_WM 0 0.0001 0.881 0.881 -0.0001 0.0001 0.697 0.697
168
CHAPTER 5
A. DISCUSSION: METHODOLOGY AND SAMPLE
As reviewed in chapter 1, the extant literature clearly shows the presence of
neuroanatomical differences between musicians and non-musicians, in a variety of
regions having functionally to do with music perception and performance, including
primary motor cortex (precentral gyrus), somatosensory cortex (postcentral gyrus),
primary auditory cortex (Heschl’s gyrus), inferior frontal gyrus, cerebellum, and corpus
callosum. Chapter 2 outlined a study undertaken to answer the question of whether or not
four years spent in conservatory-style training would be long enough to elicit noticeable
changes in these regions of interest (ROIs). The comparison group of architecture majors
was chosen for the educational similarities between architecture and music performance
degrees; previous studies in the literature have largely compared musicians to non-
musicians of many disciplines. Two main methods were used to analyze the dataset
collected: voxel-based morphometry (VBM), discussed in chapter 3, and a semi-
automated anatomical approach using the software BrainSuite, discussed in chapter 4.
VBM was part of the original analysis plan for this study; many musical experience-
dependent structural neuroplasticity studies have employed this method. However, as 11
of 52 total subjects were removed for morphological abnormalities, age at scan time, or
being architecture majors with too much musical training, it became quickly apparent that
an individual subjects-based approach was also needed.
VBM and semi-automated analysis via BrainSuite differ in three important ways.
The first of these is processing time. VBM is a completely automated method and thus
fast. BrainSuite is a semi-automated method; it involves manual editing at multiple stages
in the process, thus lengthening the amount of time to analysis completion. The second
169
way in which these two methods differ is in the comparisons that can be made. VBM is a
group analysis tool; BrainSuite allows for both individual and group analysis, meaning
that outliers can be identified and handled appropriately. The third important difference
between these methods lies in what they measure. VBM results indicate groupwise
differences in grey matter density, a measurement of the relative concentration of grey
matter within a given voxel. BrainSuite yields volumetric measures and cortical thickness
estimates. While grey matter density and grey matter volume are thought to be related,
they are distinct measurements (Ashburner & Friston, 2003).
Results from the VBM and BrainSuite analyses do not match. Indeed, the only
result present in both analyses is the left precentral gyrus increase found in architecture
freshmen as compared to architecture seniors. There are some uncorrected (for multiple
comparisons) results, which can be seen in Chapter 4, tables 6-10, that could perhaps
complement some of the VBM results presented in Chapter 3. For example, VBM
analysis yielded voxelwise differences for vocalists as compared to instrumentalists in
right POP; uncorrected results in BrainSuite showed increased TV, GM, and WM volume
in right pars orbitalis. The distance from POP to pars orbitalis seems likely larger than
can be explained away as an issue with smoothing, but, depending on extraction and bias-
field issues, this could be the same result. Similarly, the VBM result in left anterior insula
for vocalists as compared to architecture majors could be related to the uncorrected result
in left POP (TV and GM) for the same populations. Though no results in the VBM AWM
analysis matched those found in BrainSuite, an interesting uncorrected result in the latter
showed increased TV, GM, and WM in left Broca’s area for AWM students as compared
to architecture majors, adding to the left POP results, corrected for multiple comparisons,
170
present in the analysis in chapter 4. The only complementary results seen between VBM
and BrainSuite in the planned analyses were those found in left precentral gyrus,
mentioned above. The existence of these uncorrected p-value results could be due to not
just smoothing and discrepant extraction and parcellation, but also differing statistical
methods. VBM uses a permutation method to correct for multiple comparisons; perhaps
it is less conservative than the Holm-Bonferroni method used on the BrainSuite data. Use
of a less conservative statistical method might explain the high rate of false positive
results found in VBM data by several recent studies (Henley et al, 2009, Scarpazza et al,
2013). However, this line of reasoning does not explain why all results found in the
BrainSuite analyses were not also found using VBM.
VBM results in cerebellum were both highly statistically significant and sizeable,
in terms of actual number of voxels covered (Figure 1). I was unable to investigate
cerebellum in BrainSuite, but I am inclined to think that these strong grey matter density
results may well be also indicative of grey matter volume results. As this ROI is now
available in BrainSuite, it would be worth exploring in the future.
As a method, VBM is by necessity a group analysis tool. Although it has been
used more recently to compare uneven groups and even single cases against groups, these
usages of VBM have been shown to yield high rates of false positives in results
(Scarpazza et al, 2013). Smoothing has been employed successfully to help decrease the
rate of false positives (Salmond et al, 2002), but when one smooths anatomical data, the
risk of results appearing in the wrong ROI increases. Additionally, and as mentioned in
chapter 3, the subjects chosen to build the study-specific template can affect results, both
171
Figure 1: VBM results in music seniors when compared to music freshmen (p=0.014*).
in terms of location and significance. The study-specific template ideally consists of
groups with equal variances; when this is the case, VBM likely works well. However,
statistical analysis of BrainSuite data revealed that equal variances could not be assumed
between my uneven groups, and indeed, outliers existed in nearly every ROI. It is thus
possible that the small size of, and variety within, my subject population rendered the use
of VBM inadequate. Additionally, extractions and parcellation when performing any
automated structural analysis can contain mislabeling of dura and meninges as brain
tissue, or exclusion of swathes of brain matter, which can contribute to discrepant results.
MRI scanner bias-field inhomogeneities can severely decrease the reliability of
automated methods to perform extractions and parcellation, contributing to the above
issues. Only careful checking for such mishaps, and likely manual editing, can guarantee
consistency among individual brains.
Semi-automated anatomical analysis in BrainSuite is superior to automated
methods in that registration of the atlas brain to individual subject brains happens in
native space, and is constrained by anatomical landmarks. Automated usage of
BrainSuite, made possible through the software gui and script, svreg, does not yet yield
172
perfect extraction, bias-field correction, sulcal labeling, or registration. Some ROIs are
particularly susceptible to registration problems, including most subcortical regions and
the orbitofrontal cortex. However, each of the above steps is available to the user for
manual editing, including the final labeled brain volumes. With consistent editing
between brains, reliable individual subjects results can be achieved using BrainSuite,
although not yet as good as those produced by truly manual methods. Conservative
statistical methods applied to these data, accounting for unequal variances and outliers,
yield results that are much more reliable than those produced by VBM.
VBM has one major good point: voxelwise differences are displayed in color on a
template brain, making it possible to approximate location, down to a single voxel. These
results should be taken with caution of course, as the brains have been smoothed;
differences of only a few voxels located close to gyral borders could indeed be
misclassified. Semi-automated analysis via BrainSuite, while superior as a method in
many ways, yields GM and cortical thickness (as well as TV, WM, and FA) values for
entire gyri; localizing differences between brains requires the additional hand-labeling of
ROIs within those gyri. For example, if there is a difference in precentral gyrus, but the a
priori defined ROI is the hand knob, that region must be labeled separately in order to
determine if the differences in motor cortex lie there. With VBM, the location of the
results is visible immediately, without additional ROI labeling. If VBM analyses are used
as a guide for more manual methods, results in large areas like cingulate gyrus and insula
might merit the manual tracing of additional ROIs splitting those gyri up for statistical
power and a more accurate pinpointing of results. Cingulate gyrus, for example, could be
split into anterior and posterior sections; this division has functional merit as well.
173
I chose to use FSL software to run the VBM analysis, although there is one other
software package prominent in VBM studies: SPM. SPM uses random field theory, rather
than permutation theory, to analyze data (Scarpazza et al, 2013). Usage of SPM in the
music literature is now more common than usage of FSL; perhaps use of this version of
VBM methodology would yield results closer to those found in BrainSuite. Additionally,
David Shattuck has indicated that BrainSuite is capable of performing a VBM-like
analysis investigating thickness; if this can be performed for volumetric measurements,
and if the results can be localized within gyri, running this additional process in
BrainSuite would negate the only real advantage (other than time spent) that VBM
currently has over semi-automated anatomical analysis in BrainSuite.
Taking into consideration all of the above, I am inclined toward skepticism
regarding many of my VBM results, save perhaps the results in insula and cerebellum,
which I was unable to confirm in BrainSuite. It remains possible that gender differences,
remaining issues with bias-field correction, and/or registration inconsistencies –
particularly in prefrontal cortex and temporal pole – are affecting the results. I thus
hesitantly conclude from this study that: 1. Music majors do exhibit morphological
differences when compared to architecture majors, particularly in left Broca’s area of the
inferior frontal gyrus, and in grey matter density in left anterior insula. 2. These
differences in Broca’s area appear to be driven by music freshmen and instrumentalists.
3. Music seniors exhibit increased TV, GM, and WM in left postcentral gyrus, and
increased grey matter density in right cerebellum, as compared to music freshmen.
4. Architecture seniors exhibit increased WM in right Broca’s area as compared to
architecture freshmen. 5. There are morphological differences between instrument
174
groups, in both precentral (GM) and postcentral gyri (FA). 6. The AWM subjects do not
represent a clean middle group between architecture majors and music majors. 7. Age of
onset of primary instrument musical training is likely more relevant than the age of onset
of all musical training, and is negatively correlated with WM volume in left insula and
FA in right precentral gyrus in all music majors, with GM volume in left precentral gyrus
in instrumentalists, and with TV, GM volume, WM volume, and FA in left PTR, CT in
left Heschl’s gyrus, and FA in both right and left cingulate gyri, in vocalists.
B. SUMMARY
In summary, the most relevant results of this study are as follows:
1. To answer my first research question, I compared the brains of music seniors to music
freshmen. In the seniors, I found increased grey matter density in right cerebellum, and
total volume, grey matter volume, and white matter volume increases in left postcentral
gyrus (Figures 1-2).
175
Figure 2: BrainSuite results in music seniors when compared to music freshmen in left postcentral gyrus.
2. To answer my second research question, I compared the brains of music majors to
architecture majors. In the music majors, I found increased grey matter density in left
anterior insula, and total volume and grey matter volume increases, as well as a statistical
trend toward increased white matter volume, in left Broca’s area (Figures 3-4).
176
Figure 3: VBM results in music majors when compared to architecture majors (p=0.0362*).
Figure 4: BrainSuite results in music majors when compared to architecture majors in left Broca’s area.
177
3. I investigated the possibility of an effect of instrumentation in the music major
population by splitting the group into piano majors (n=6), strings majors (n=9), and voice
majors (n=10). I found that piano majors exhibit a trend toward increased grey matter
volume in left precentral gyrus when compared to strings majors (Figure 5), and, when
compared to voice majors, increased fractional anisotropy in right postcentral gyrus, and
a trend toward increased fractional anisotropy in right precentral gyrus (Figure 6).
Figure 5: BrainSuite results in piano majors when compared to strings majors in left precentral gyrus.
Figure 6: BrainSuite results in piano majors when compared to voice majors in right precentral gyrus and
right postcentral gyrus.
178
4. Finally, I investigated the possibility of an effect of the age of onset of musical training
separately in these three instrument groups. I found that instrumentalists who start
learning their primary instrument at an earlier age have larger grey matter volume in left
precentral gyrus (Figure 7). Vocalists who start their vocal training at an earlier age have
larger total volume, and white matter volume, and statistical trends toward larger grey
matter volume and fractional anisotropy in left pars triangularis, as well as increased
cortical thickness in left Heschl’s gyrus, and larger fractional anisotropy values bilaterally
in cingulate gyrus (Figure 8).
Figure 7: BrainSuite correlation results in instrumentalists in left precentral gyrus.
179
Figure 8: BrainSuite correlation results in vocalists in left pars triangularis, left Heschl’s gyrus, and
bilateral cingulate gyrus.
180
C. FUTURE DIRECTIONS
Many things could improve the results presented here. Within the parameters of
the current study, it would first be useful to employ the new bias-field correction tool and
manually edit prefrontal and temporal pole ROIs in particular, to see if there are
differences in those regions between groups, particularly as orbitofrontal cortex has been
linked in terms of functional usage to architects and aesthetic appreciation (Kirk et al,
2009). Second, as inclusion of the cerebellum is now possible, this region should be
investigated. Third, trying SPM and BrainSuite’s VBM-style analysis could help
elucidate the validity of VBM as a method, and the latter, performed in BrainSuite, could
provide coordinate information for results within ROIs. Fourth, diffusion values other
than FA should be investigated, to paint a clearer picture of differences in white matter
microstructure. Along with the additional values, a method allowing for investigation of
all white matter regions, particularly corpus callosum and internal capsule, should be
used. Fifth, possible gender differences should be explored within the BrainSuite results.
Additionally, the recruitment of more subjects would help with statistical power, in all of
these analyses.
To add to this study, it would be interesting to see if the architecture majors are
representative of current non-music majors at this university. The architecture majors
may be too similar to the music majors; in order to truly answer the question of whether
or not differences exist between conservatory musicians and non-musicians in the
predicted areas, a third group of students should be investigated. These could be
humanities majors, or a mixed group of university students with minimal musical
backgrounds. It would also be interesting to see how conservatory arts majors (both
181
music and architecture) compare to their professional counterparts, either via longitudinal
or cross-sectional study. Regarding longitudinal study, as the freshmen scanned for this
project are now seniors, it would be interesting to see how their brains have changed over
the course of four years of conservatory-style training.
I intended to add to the extant literature by including both genders and voice
majors. This latter choice may have entirely changed the results I would have gotten, had
I stuck to instrumentalists. It would be interesting to look at brass players, specifically
trumpet and tuba players, whose instruments require the opposite asymmetrical hand
usage to strings players; these brass players use the right hand for dexterous movement
and hold the instrument with the left hand. This could very well yield an opposite
asymmetry in both precentral and postcentral gyri. The vocalists themselves would be
interesting to investigate, particularly since they have a lot of regional negative
correlations with primary instrument age of onset. Any of these additional comparisons
would help advance our understanding of how the experience of musical training can
affect structural neuroplasticity, the study of which holds several important implications.
D. IMPLICATIONS
The possible benefits of plasticity in the human brain should be kept in mind
when evaluating the studies addressing structural brain differences in musicians
compared to non-musicians. According to the framework put forth by Lövdén (Lövdén,
et al 2011), structural neuroplasticity often represents an increase of flexibility in use in a
given region. This flexibility may translate to increased support for cognitive transfer
effects, which have been demonstrated in a wide variety of disciplines in relation to
182
Figure 9: The two most immediately relevant implications of musical experience-dependent neuroplasticity
research.
musical training, including linguistic skills (including language acquisition and reading),
mathematics, auditory skills, multiple types of intelligence, academic achievement
(usually measured through standardized testing), creativity, pro-social behavior, and
mental and physical wellness (Hallam, 2010, Kraus, 2010, Masutani et al, 2010,
Milanovev and Tervaniemi, 2011, Corrigall et al, 2013, Ferreri et al, 2013, Tierney &
Kraus, 2013). Musical training has been demonstrated to have a positive effect on
linguistic abilities in children with developmental disabilities, such as dyslexia (Hallam,
2010). And, as discussed in Wan and Schlaug’s 2010 review, the ability of the brain to
change in response to musical training even late into adulthood strengthens the viability
of musical therapy as an intervention in cases of neuroaffective disorders. The
183
implications of musical experience-dependent structural neuroplasticity thus run the
gamut from quality of life improvements to academic gains, from childhood to late
adulthood, and from healthy individuals to those suffering from various neuroaffective
disorders, though the most immediately relevant applications of this field can be found in
education and music therapy (Figure 9). The field, reviewed in chapter 1, is too important
to leave in a state of such mixed results; In addition to the future directions proposed
above for this study, further studies addressing both genders in larger groups of highly
skilled musicians using different instruments (vocalists, composers, and conductors
included) must be conducted in order to draw truly generalizable conclusions about how
musicians’ brains may differ from non-musicians’ brains, and how these differences
might affect cognitive transfer effects and musical therapy.
184
REFERENCES
Abdul-Kareem IA, Stancak A, Parkes LM, Al-Ameen M, Alghamdi J, Aldhafeeri FM,
Embleton K, Morris D, Sluming V (2011) Plasticity of the superior and middle
cerebellar peduncles in musicians revealed by quantitative analysis of volume and
number of streamlines based on diffusion tensor tractography. Cerebellum
10:611-623.
Abdul-Kareem IA, Stancak A, Parkes LM, Sluming V (2011) Increased gray matter
volume of left pars opercularis in male orchestral musicians correlate positively
with years of musical performance. J Magn Reson Imaging 33:24-32.
Alba-Ferrara LM, de Erausquin GA (2013) What does anisotropy measure? Insights from
increased and decreased anisotropy in selective fiber tracts in schizophrenia. Front
Integr Neurosci 7:9.
Alexander AL, Lee JE, Lazar M, Field AS (2007) Diffusion tensor imaging of the brain.
Neurotherapeutics 4:316-329.
Allen JS, Bruss J, Brown CK, Damasio H (2005) Normal neuroanatomical variation due
to age: the major lobes and a parcellation of the temporal region. Neurobiol Aging
26:1245-1260; discussion 1279-1282.
Allen JS, Emmorey K, Bruss J, Damasio H (2007) An MR volumetric study of the motor
hand region and insula in deaf, hearing, and hearing-signing individuals. Poster:
Cognitive Neuroscience Society.
Altenmüller E, Kesselring J, Wiesendanger M (2006) Music, motor control and the brain.
Oxford; New York: Oxford University Press.
185
Amunts K, Jancke L, Mohlberg H, Steinmetz H, Zilles K (2000) Interhemispheric
asymmetry of the human motor cortex related to handedness and gender.
Neuropsychologia 38:304-312.
Amunts K, Schlaug G, Jancke L, Steinmetz H, Schleicher A, Dabringhaus A, Zilles K
(1997) Motor cortex and hand motor skills: structural compliance in the human
brain. Hum Brain Mapp 5:206-215.
Amunts K, Schlaug G, Schleicher A, Steinmetz H, Dabringhaus A, Roland PE, Zilles K
(1996) Asymmetry in the human motor cortex and handedness. Neuroimage
4:216-222.
Andersson JLR, Jenkinson M, Smith S (2007a) Non-linear optimization. FMRIB
technical report TR07JA1 from www.fmrib.ox.ac.uk/analysis/techrep
Andersson JLR, Jenkinson M, Smith S (2007b) Non-linear registration, aka spatial
normalization. FMRIB technical report TR07JA2 from
www.fmrib.ox.ac.uk/analysis/techrep
Andreasen NC, Ramchandran K (2012) Creativity in art and science: are there two
cultures? Dialogues Clin Neurosci 14:49-54.
Ashburner J, Friston KJ (2000) Voxel-based morphometry--the methods. Neuroimage
11:805-821.
Aydin K, Ciftci K, Terzibasioglu E, Ozkan M, Demirtas A, Sencer S, Minareci O (2005)
Quantitative proton MR spectroscopic findings of cortical reorganization in the
auditory cortex of musicians. AJNR Am J Neuroradiol 26:128-136.
Aydin K, Ucar A, Oguz KK, Okur OO, Agayev A, Unal Z, Yilmaz S, Ozturk C (2007)
Increased gray matter density in the parietal cortex of mathematicians: a voxel-
186
based morphometry study. AJNR Am J Neuroradiol 28:1859-1864.
Bangert M, Schlaug G (2006) Specialization of the specialized in features of external
human brain morphology. Eur J Neurosci 24:1832-1834.
Bengtsson SL, Nagy Z, Skare S, Forsman L, Forssberg H, Ullen F (2005) Extensive
piano practicing has regionally specific effects on white matter development. Nat
Neurosci 8:1148-1150.
Bermudez P, Zatorre RJ (2005) Differences in gray matter between musicians and non-
musicians. Ann N Y Acad Sci 1060:395-399.
Bharucha JJ, Curtis M, Paroo K (2006) Varieties of musical experience. Cognition
100:131-172.
Bhushan C, Haldar JP, Joshi AA, Leahy RM (2012) Correcting susceptibility-induced
distortion in diffusion-weighted MRI using constrained nonrigid registration.
Asia-Pacific Signal & Information Processing Association Annual Summit and
Conference (APSIPA ASC):1-9.
Black JE, Isaacs KR, Anderson BJ, Alcantara AA, Greenough WT (1990) Learning
causes synaptogenesis, whereas motor activity causes angiogenesis, in cerebellar
cortex of adult rats. Proc Natl Acad Sci U S A 87:5568-5572.
Boyke J, Driemeyer J, Gaser C, Buchel C, May A (2008) Training-induced brain
structure changes in the elderly. J Neurosci 28:7031-7035.
Brang D, Ramachandran VS (2011) Survival of the synaesthesia gene: why do people
hear colors and taste words? PLoS One 11:e1001205.
Brattico E, Bogert B, Jacobsen T (2013) Toward a neural chronometry for the aesthetic
experience of music. Front Psychol 4:206.
187
Brown S (2000) Evolutionary models of music: from sexual selection to group selection.
In Thompson NS, Tonneau F (Eds.), Perspectives in Ethology 13: Evolution,
Culture, and Behavior. Plenum, New York. 221-281.
Chakravarty A (2012) The neural circuitry of visual artistic production and appreciation:
A proposition. Ann Indian Acad Neurol 15:71-75.
Clos M, Amunts K, Laird AR, Fox PT, Eickhoff SB (2013) Tackling the multifunctional
nature of Broca's region meta-analytically: co-activation-based parcellation of
area 44. Neuroimage 83:174-188.
Corrigall KA, Schellenberg EG, Misura NM (2013) Music training, cognition, and
personality. Front Psychol 4:222.
Craig AD (2009) How do you feel--now? The anterior insula and human awareness. Nat
Rev Neurosci 10:59-70.
Cross I (2001) Music, cognition, culture, and evolution. Ann N Y Acad Sci 930:28-42.
Damasio AR, Grabowski TJ, Bechara A, Damasio H, Ponto LL, Parvizi J, Hichwa RD
(2000) Subcortical and cortical brain activity during the feeling of self-generated
emotions. Nat Neurosci 3:1049-1056.
Dehaene S, Spelke E, Pinel P, Stanescu R, Tsivkin S (1999) Sources of mathematical
thinking: behavioral and brain-imaging evidence. Science 284:970-974.
Deutsch D, Henthorn T, Marvin E, Xu H (2006) Absolute pitch among American and
Chinese conservatory students: prevalence differences, and evidence for a speech-
related critical period. J Acoust Soc Am 119:719-722.
Dissanayake E (2000). Antecedents of the temporal arts in early mother-infant
interactions. In Wallin N, Merker B, Brown S (Eds.), The Origins of Music. MIT
188
Press, Cambridge, Massachusetts. 389-407.
Draganski B, Gaser C, Busch V, Schuierer G, Bogdahn U, May A (2004)
Neuroplasticity: changes in grey matter induced by training. Nature 427:311-312.
Draganski B, Gaser C, Kempermann G, Kuhn HG, Winkler J, Buchel C, May A (2006)
Temporal and spatial dynamics of brain structure changes during extensive
learning. J Neurosci 26:6314-6317.
Driemeyer J, Boyke J, Gaser C, Buchel C, May A (2008) Changes in gray matter induced
by learning--revisited. PLoS One 3:e2669.
Elbert T, Pantev C, Wienbruch C, Rockstroh B, Taub E (1995) Increased cortical
representation of the fingers of the left hand in string players. Science 270:305-
307.
Fairhall SL, Ishai A (2008) Neural correlates of object indeterminacy in art compositions.
Conscious Cogn 17:923-932.
Ferreri L, Aucouturier JJ, Muthalib M, Bigand E, Bugaiska A (2013) Music improves
verbal memory encoding while decreasing prefrontal cortex activity: an fNIRS
study. Front Hum Neurosci 7:779.
Garthe A, Kempermann G (2013) An old test for new neurons: refining the Morris water
maze to study the functional relevance of adult hippocampal neurogenesis. Front
Neurosci 7:63.
Gaser C, Schlaug G (2003) Brain structures differ between musicians and non-musicians.
J Neurosci 23:9240-9245.
Giedd JN (2008) The teen brain: insights from neuroimaging. J Adolesc Health 42:335-
343.
189
Giedd JN, Rapoport JL (2010) Structural MRI of pediatric brain development: what have
we learned and where are we going? Neuron 67:728-734.
Giedd JN, Vaituzis AC, Hamburger SD, Lange N, Rajapakse JC, Kaysen D, Vauss YC,
Rapoport JL (1996) Quantitative MRI of the temporal lobe, amygdala, and
hippocampus in normal human development: ages 4-18 years. J Comp Neurol
366:223-230.
Gogtay N, Giedd JN, Lusk L, Hayashi KM, Greenstein D, Vaituzis AC, Nugent TF, 3rd,
Herman DH, Clasen LS, Toga AW, Rapoport JL, Thompson PM (2004) Dynamic
mapping of human cortical development during childhood through early
adulthood. Proc Natl Acad Sci U S A 101:8174-8179.
Good CD, Johnsrude IS, Ashburner J, Henson RN, Friston KJ, Frackowiak RS (2001) A
voxel-based morphometric study of ageing in 465 normal adult human brains.
Neuroimage 14:21-36.
Grabner RH, Ansari D, Reishofer G, Stern E, Ebner F, Neuper C (2007) Individual
differences in mathematical competence predict parietal brain activation during
mental calculation. Neuroimage 38:346-356.
Groussard M, La Joie R, Rauchs G, Landeau B, Chetelat G, Viader F, Desgranges B,
Eustache F, Platel H (2010) When music and long-term memory interact: effects
of musical expertise on functional and structural plasticity in the hippocampus.
PLoS One 5.
Hallam S (2010) 21
st
century conceptions of musical ability. Psychology of Music
38:308-331.
Halpern AR, Zatorre RJ (1999) When that tune runs through your head: a PET
190
investigation of auditory imagery for familiar melodies. Cereb Cortex 9:697-704.
Han Y, Yang H, Lv YT, Zhu CZ, He Y, Tang HH, Gong QY, Luo YJ, Zang YF, Dong Q
(2009) Gray matter density and white matter integrity in pianists' brain: a
combined structural and diffusion tensor MRI study. Neurosci Lett 459:3-6.
Hänggi J, Koeneke S, Bezzola L, Jäncke L (2010) Structural neuroplasticity in the
sensorimotor network of professional female ballet dancers. Hum Brain Mapp
31:1196-1206.
Henley SM, Ridgway GR, Scahill RI, Kloppel S, Tabrizi SJ, Fox NC, Kassubek J (2009)
Pitfalls in the use of voxel-based morphometry as a biomarker: examples from
huntington disease. AJNR Am J Neuroradiol 31:711-719.
Hihara S, Notoya T, Tanaka M, Ichinose S, Ojima H, Obayashi S, Fujii N, Iriki A (2006)
Extension of corticocortical afferents into the anterior bank of the intraparietal
sulcus by tool-use training in adult monkeys. Neuropsychologia 44:2636-2646.
Howard-Jones PA, Blakemore SJ, Samuel EA, Summers IR, Claxton G (2005) Semantic
divergence and creative story generation: an fMRI investigation. Brain Res Cogn
Brain Res 25:240-250.
Hubbard E, Ramachandran VS (2003) The phenomenology of synaesthesia. J Conscious
Stud 10:49-57.
Hugenschmidt CE, Peiffer AM, Kraft RA, Casanova R, Deibler AR, Burdette JH,
Maldjian JA, Laurienti PJ (2008) Relating imaging indices of white matter
integrity and volume in healthy older adults. Cereb Cortex 18:433-442.
Huron D (2001) Is music an evolutionary adaptation? Ann N Y Acad Sci 930:43-61.
Hutchinson S, Lee LH, Gaab N, Schlaug G (2003) Cerebellar volume of musicians.
191
Cereb Cortex 13:943-949.
Hyde KL, Lerch J, Norton A, Forgeard M, Winner E, Evans AC, Schlaug G (2009)
Musical training shapes structural brain development. J Neurosci 29:3019-3025.
Imfeld A, Oechslin MS, Meyer M, Loenneker T, Jancke L (2009) White matter plasticity
in the corticospinal tract of musicians: a diffusion tensor imaging study.
Neuroimage 46:600-607.
Jacobsen T (2010) Beauty and the brain: culture, history and individual differences in
aesthetic appreciation. J Anat 216:184-191.
Jacobsen T, Schubotz RI, Hofel L, Cramon DY (2006) Brain correlates of aesthetic
judgment of beauty. Neuroimage 29:276-285.
James CE, Oechslin MS, Van De Ville D, Hauert CA, Descloux C, Lazeyras F (2013)
Musical training intensity yields opposite effects on grey matter density in
cognitive versus sensorimotor networks. Brain Struct Funct.
Jäncke L, Koeneke S, Hoppe A, Rominger C, Hanggi J (2009) The architecture of the
golfer's brain. PLoS One 4:e4785.
Jenkinson M, Bannister P, Brady M, Smith S (2002) Improved optimization for the
robust and accurate linear registration and motion correction of brain images.
Neuroimage 17:825-841.
Jenkinson M, Smith S (2001) A global optimisation method for robust affine registration
of brain images. Med Image Anal 5:143-156.
Jones DK, Knosche TR, Turner R (2013) White matter integrity, fiber count, and other
fallacies: the do's and don'ts of diffusion MRI. Neuroimage 73:239-254.
Jones TA (1999) Multiple synapse formation in the motor cortex opposite unilateral
192
sensorimotor cortex lesions in adult rats. J Comp Neurol 414:57-66.
Joshi AA, Pantazis D, Li Q, Damasio H, Shattuck DW, Toga AW, Leahy RM (2010)
Sulcal set optimization for cortical surface registration. Neuroimage 50:950-959.
Joshi AA, Shattuck DW, Leahy RM (2012) A method for automated cortical surface
registration and labeling. Lect Notes Comput Sc 7359:180-189.
Kahnt T, Chang LJ, Park SQ, Heinzle J, Haynes JD (2012) Connectivity-based
parcellation of the human orbitofrontal cortex. J Neurosci 32:6240-6250.
Keenan JP, Thangaraj V, Halpern AR, Schlaug G (2001) Absolute pitch and planum
temporale. Neuroimage 14:1402-1408.
Kempermann G, Gast D, Gage FH (2002) Neuroplasticity in old age: sustained fivefold
induction of hippocampal neurogenesis by long-term environmental enrichment.
Ann Neurol 52:135-143.
Kempermann G, Kuhn HG, Gage FH (1997) More hippocampal neurons in adult mice
living in an enriched environment. Nature 386:493-495.
Kempermann G, Kuhn HG, Gage FH (1998) Experience-induced neurogenesis in the
senescent dentate gyrus. J Neurosci 18:3206-3212.
Kempermann G, Wiskott L, Gage FH (2004) Functional significance of adult
neurogenesis. Curr Opin Neurobiol 14:186-191.
Kim JH, Haldar J, Liang ZP, Song SK (2009) Diffusion tensor imaging of mouse brain
stem and cervical spinal cord. J Neurosci Methods 176:186-191.
Kirk U, Skov M, Christensen MS, Nygaard N (2009) Brain correlates of aesthetic
expertise: a parametric fMRI study. Brain Cogn 69:306-315.
Kleim JA, Hogg TM, VandenBerg PM, Cooper NR, Bruneau R, Remple M (2004)
193
Cortical synaptogenesis and motor map reorganization occur during late, but not
early, phase of motor skill learning. J Neurosci 24:628-633.
Koelsch S, Fritz T, DY VC, Muller K, Friederici AD (2006) Investigating emotion with
music: an fMRI study. Hum Brain Mapp 27:239-250.
Koelsch S, Gunter TC, Wittfoth M, Sammler D (2005) Interaction between syntax
processing in language and in music: an ERP Study. J Cogn Neurosci 17:1565-
1577.
Koelsch S, Siebel WA (2005) Towards a neural basis of music perception. Trends Cogn
Sci 9:578-584.
Kowatari Y, Lee SH, Yamamura H, Nagamori Y, Levy P, Yamane S, Yamamoto M
(2009) Neural networks involved in artistic creativity. Hum Brain Mapp 30:1678-
1690.
Kraus N, Chandrasekaran B (2010) Music training for the development of auditory skills.
Nat Rev Neurosci 11:599-605.
Lee DJ, Chen Y, Schlaug G (2003) Corpus callosum: musician and gender effects.
Neuroreport 14:205-209.
Lentini E, Kasahara M, Arver S, Savic I (2012) Sex differences in the human brain and
the impact of sex chromosomes and sex hormones. Cereb Cortex 23:2322-2336.
Leow AD, Zhan L, Zhu S, Hageman N, Chiang M-C, Barysheva M, Toga AW,
McMahon KL, de Zubicaray GI, Wright MJ, Thompson PM (2009) White matter
integrity measured by fractional anisotropy correlates poorly with actual
individual fiber anisotropy. Biomedical Imaging: From Nano to Macro. ISBI '09.
IEEE International Symposium:622-625.
194
Loui P, Li HC, Hohmann A, Schlaug G (2010) Enhanced cortical connectivity in absolute
pitch musicians: a model for local hyperconnectivity. J Cogn Neurosci 23:1015-
1026.
Lövdén M, Backman L, Lindenberger U, Schaefer S, Schmiedek F (2010) A theoretical
framework for the study of adult cognitive plasticity. Psychol Bull 136:659-676.
Maess B, Koelsch S, Gunter TC, Friederici AD (2001) Musical syntax is processed in
Broca's area: an MEG study. Nat Neurosci 4:540-545.
Maguire EA, Gadian DG, Johnsrude IS, Good CD, Ashburner J, Frackowiak RS, Frith
CD (2000) Navigation-related structural change in the hippocampi of taxi drivers.
Proc Natl Acad Sci U S A 97:4398-4403.
Maguire EA, Spiers HJ, Good CD, Hartley T, Frackowiak RS, Burgess N (2003)
Navigation expertise and the human hippocampus: a structural brain imaging
analysis. Hippocampus 13:250-259.
Mann SL, Hazlett EA, Byne W, Hof PR, Buchsbaum MS, Cohen BH, Goldstein KE,
Haznedar MM, Mitsis EM, Siever LJ, Chu KW (2011) Anterior and posterior
cingulate cortex volume in healthy adults: effects of aging and gender differences.
Brain Res 1401:18-29.
Mason MF, Norton MI, Van Horn JD, Wegner DM, Grafton ST, Macrae CN (2007)
Wandering minds: the default network and stimulus-independent thought. Science
315:393-395.
Masutani T, Yamamoto Y, Konishi J, Maeda K (2010) Effects of music and art education
in early life and oral functions on the QOL of the Takarazuka Revue Company
195
OG compared with general elderly females. Psychogeriatrics 10:4-14.
May A, Hajak G, Ganssbauer S, Steffens T, Langguth B, Kleinjung T, Eichhammer P
(2007) Structural brain alterations following 5 days of intervention: dynamic
aspects of neuroplasticity. Cereb Cortex 17:205-210.
Mechelli A, Crinion JT, Noppeney U, O'Doherty J, Ashburner J, Frackowiak RS, Price
CJ (2004) Neurolinguistics: structural plasticity in the bilingual brain. Nature
431:757.
Milovanov R, Tervaniemi M (2011) The Interplay between Musical and Linguistic
Aptitudes: A Review. Front Psychol 2:321.
Mithen SJ (2006) The singing neanderthals : the origins of music, language, mind, and
body. Cambridge, Mass.: Harvard University Press.
Molnar-Szakacs I, Overy K (2006) Music and mirror neurons: from motion to 'e'motion.
Soc Cogn Affect Neurosci 1:235-241.
Mutschler I, Reinbold C, Wankerl J, Seifritz E, Ball T (2013) Structural basis of empathy
and the domain general region in the anterior insular cortex. Front Hum Neurosci
7:177.
Nadal M, Munar E, Capo MA, Rossello J, Cela-Conde CJ (2008) Towards a framework
for the study of the neural correlates of aesthetic preference. Spat Vis 21:379-396.
Parsons LM (2001) Exploring the functional neuroanatomy of music performance,
perception, and comprehension. Ann N Y Acad Sci 930:211-231.
Peretz I, Zatorre RJ (2003) The cognitive neuroscience of music. Oxford ; New York:
Oxford University Press.
Phan KL, Wager T, Taylor SF, Liberzon I (2002) Functional neuroanatomy of emotion: a
196
meta-analysis of emotion activation studies in PET and fMRI. Neuroimage
16:331-348.
Platel H, Price C, Baron JC, Wise R, Lambert J, Frackowiak RS, Lechevalier B, Eustache
F (1997) The structural components of music perception. A functional anatomical
study. Brain 120 ( Pt 2):229-243.
Qazi AA, Radmanesh A, O'Donnell L, Kindlmann G, Peled S, Whalen S, Westin CF,
Golby AJ (2009) Resolving crossings in the corticospinal tract by two-tensor
streamline tractography: Method and clinical assessment using fMRI.
Neuroimage 47 Suppl 2:T98-106.
Rueckert D, Sonoda LI, Hayes C, Hill DL, Leach MO, Hawkes DJ (1999) Nonrigid
registration using free-form deformations: application to breast MR images. IEEE
Trans Med Imaging 18:712-721.
Sakai K, Hikosaka O, Miyauchi S, Takino R, Tamada T, Iwata NK, Nielsen M (1999)
Neural representation of a rhythm depends on its interval ratio. J Neurosci
19:10074-10081.
Salmond CH, Ashburner J, Vargha-Khadem F, Connelly A, Gadian DG, Friston KJ
(2002) Distributional assumptions in voxel-based morphometry. Neuroimage
17:1027-1030.
Sarnthein J, vonStein A, Rappelsberger P, Petsche H, Rauscher FH, Shaw GL (1997)
Persistent patterns of brain activity: an EEG coherence study of the positive effect
of music on spatial-temporal reasoning. Neurol Res 19:107-116.
Scarpazza C, Sartori G, De Simone MS, Mechelli A (2013) When the single matters
more than the group: very high false positive rates in single case Voxel Based
197
Morphometry. Neuroimage 70:175-188.
Schellenberg EG (2003) Does exposure to music have beneficial side effects? In Peretz I,
Zatorre R (Eds.), The cognitive neuroscience of music. Oxford University Press,
Oxford.
Schlaug G (2006) Brain structures of musicians: executive functions and morphological
implications. In Altenmüller E, Wiesendanger M, Kesselring J (Eds.), Music,
Motor Control, and The Brain. Oxford University Press, New York.
Schlaug G, Jäncke L, Huang Y, Staiger JF, Steinmetz H (1995) Increased corpus
callosum size in musicians. Neuropsychologia 33:1047-1055.
Schlaug G, Norton A, Overy K, Winner E (2005) Effects of music training on the child's
brain and cognitive development. Ann N Y Acad Sci 1060:219-230.
Schmithorst VJ, Wilke M (2002) Differences in white matter architecture between
musicians and non-musicians: a diffusion tensor imaging study. Neurosci Lett
321:57-60.
Schneider P, Scherg M, Dosch HG, Specht HJ, Gutschalk A, Rupp A (2002) Morphology
of Heschl's gyrus reflects enhanced activation in the auditory cortex of musicians.
Nat Neurosci 5:688-694.
Schneider P, Sluming V, Roberts N, Bleeck S, Rupp A (2005a) Structural, functional,
and perceptual differences in Heschl's gyrus and musical instrument preference.
Ann N Y Acad Sci 1060:387-394.
Schneider P, Sluming V, Roberts N, Scherg M, Goebel R, Specht HJ, Dosch HG, Bleeck
S, Stippich C, Rupp A (2005b) Structural and functional asymmetry of lateral
Heschl's gyrus reflects pitch perception preference. Nat Neurosci 8:1241-1247.
198
Scholz J, Klein MC, Behrens TE, Johansen-Berg H (2009) Training induces changes in
white-matter architecture. Nat Neurosci 12:1370-1371.
Sergent J, Zuck E, Terriah S, MacDonald B (1992) Distributed neural network
underlying musical sight-reading and keyboard performance. Science 257:106-
109.
Sergent J, Zuck E, Terriah S, MacDonald B (1992) Distributed neural network
underlying musical sight-reading and keyboard performance. Science 257:106-
109.
Shattuck DW, Joshi AA, Pantazis D, Kan E, Dutton RA, Sowell ER, Thompson PM,
Toga AW, Leahy RM (2009) Semi-automated method for delineation of
landmarks on models of the cerebral cortex. J Neurosci Methods 178:385-392.
Shattuck DW, Leahy RM (2002) BrainSuite: an automated cortical surface identification
tool. Med Image Anal 6:129-142.
Sluming V, Barrick T, Howard M, Cezayirli E, Mayes A, Roberts N (2002) Voxel-based
morphometry reveals increased gray matter density in Broca's area in male
symphony orchestra musicians. Neuroimage 17:1613-1622.
Smith SM (2002) Fast robust automated brain extraction. Hum Brain Mapp 17:143-155.
Smith SM, Jenkinson M, Woolrich MW, Beckmann CF, Behrens TE, Johansen-Berg H,
Bannister PR, De Luca M, Drobnjak I, Flitney DE, Niazy RK, Saunders J,
Vickers J, Zhang Y, De Stefano N, Brady JM, Matthews PM (2004) Advances in
functional and structural MR image analysis and implementation as FSL.
Neuroimage 23 Suppl 1:S208-219.
Soares JM, Marques P, Alves V, Sousa N (2013) A hitchhiker's guide to diffusion tensor
199
imaging. Front Neurosci 7:31.
Song H, Kempermann G, Overstreet Wadiche L, Zhao C, Schinder AF, Bischofberger J
(2005) New neurons in the adult mammalian brain: synaptogenesis and functional
integration. J Neurosci 25:10366-10368.
Steele CJ, Bailey JA, Zatorre RJ, Penhune VB (2013) Early musical training and white-
matter plasticity in the corpus callosum: evidence for a sensitive period. J
Neurosci 33:1282-1290.
Terrazas A, McNaughton BL (2000) Brain growth and the cognitive map. Proc Natl Acad
Sci U S A 97:4414-4416.
Thompson PM, Sowell ER, Gogtay N, Giedd JN, Vidal CN, Hayashi KM, Leow A,
Nicolson R, Rapoport JL, Toga AW (2005) Structural MRI and brain
development. Int Rev Neurobiol 67:285-323.
Tierney A, Kraus N (2013) Music training for the development of reading skills. Prog
Brain Res 207:209-241.
Trachtenberg JT, Chen BE, Knott GW, Feng G, Sanes JR, Welker E, Svoboda K (2002)
Long-term in vivo imaging of experience-dependent synaptic plasticity in adult
cortex. Nature 420:788-794.
Trainor LJ (2005) Are there critical periods for musical development? Dev Psychobiol
46(3):262-78.
Ukkola-Vuoti L, Kanduri C, Oikkonen J, Buck G, Blancher C, Raijas P, Karma K,
Lahdesmaki H, Jarvela I (2013) Genome-wide copy number variation analysis in
extended families and unrelated individuals characterized for musical aptitude and
creativity in music. PLoS One 8:e56356.
200
Vaidya JG, Paradiso S, Boles Ponto LL, McCormick LM, Robinson RG (2007) Aging,
grey matter, and blood flow in the anterior cingulate cortex. Neuroimage 37:1346-
1353.
Wan CY, Schlaug G (2010) Music making as a tool for promoting brain plasticity across
the life span. Neuroscientist 16:566-577.
Wei G, Zhang Y, Jiang T, Luo J (2011) Increased cortical thickness in sports experts: a
comparison of diving players with the controls. PLoS One 6:e17112.
Whitford TJ, Rennie CJ, Grieve SM, Clark CR, Gordon E, Williams LM (2007) Brain
maturation in adolescence: concurrent changes in neuroanatomy and
neurophysiology. Hum Brain Mapp 28:228-237.
Zarate JM (2013) The neural control of singing. Front Hum Neurosci 7:237.
Zatorre RJ (2003) Absolute pitch: a model for understanding the influence of genes and
development on neural and cognitive function. Nat Neurosci 6:692-695.
Zatorre RJ, Fields RD, Johansen-Berg H (2012) Plasticity in gray and white:
neuroimaging changes in brain structure during learning. Nat Neurosci 15:528-
536.
Zhang Y, Brady M, Smith S (2001) Segmentation of brain MR images through a hidden
Markov random field model and the expectation-maximization algorithm. IEEE
Trans Med Imaging 20:45-57.
Abstract (if available)
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Mapping the neural architecture of concepts
PDF
Behavioral and neural correlates of core-self and autobiographical-self processes
PDF
Contribution of fundamental brain networks to individual differences in social cognition
PDF
The brain's virtuous cycle: an investigation of gratitude and good human conduct
PDF
Psychopathic traits and the fronto-amygdala circuit in a community sample: the role of trait anxiety
PDF
Role of emotion and empathy in moral judgments of real-life situations
PDF
Individual differences in heart rate response and expressive behavior during social emotions: effects of resting cardiac vagal tone and culture, and relation to the default mode network
PDF
The role of music training on behavioral and neurophysiological indices of speech-in-noise perception: a meta-analysis and randomized-control trial
PDF
The neuroscience of ambivalent and ambiguous feelings
PDF
Engagement of the action observation network through functional magnetic resonance imaging with implications for stroke rehabilitation
PDF
Experience-dependent neuroplasticity of the dorsal striatum and prefrontal cortex in the MPTP-lesioned mouse model of Parkinson’s disease
PDF
Regional cerebral hemodynamic responses to hypoxia in humans using MRI and NIRS
PDF
Mapping multi-scale connectivity of the mouse posterior parietal cortex
PDF
Brain and behavior correlates of intrinsic motivation and skill learning
PDF
Functional properties of the superficial cortical interneurons
PDF
Implementing competency based education training in a technical and vocational institute in Brunei: a gap analysis
PDF
Dedifferentiation of emotion regulation strategies in the aging brain: an MVPA investigation
PDF
Refined gold from along the Royal Highway: musical training at the California Missions as a model for present-day pedagogy
PDF
Musculotendinous adaptations associated with a functional training program: the transition to barefoot running
PDF
A curriculum model for instrumental music education in the schools of Queensland, Australia
Asset Metadata
Creator
Tuttle, Meghen Miles
(author)
Core Title
Majoring in music: how conservatory training changes the brain
School
College of Letters, Arts and Sciences
Degree
Doctor of Philosophy
Degree Program
Neuroscience
Publication Date
03/14/2014
Defense Date
01/14/2014
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
architect,Architecture,BrainSuite,conservatory,MRI,music,musician,neuroanatomy,neuroplasticity,OAI-PMH Harvest,Training,VBM
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Damasio, Antonio (
committee chair
), Damasio, Hanna (
committee chair
), Baker, Laura A. (
committee member
), Cutietta, Robert (
committee member
)
Creator Email
meghen@alumni.usc.edu,uscmeg@gmail.com
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c3-369633
Unique identifier
UC11296395
Identifier
etd-TuttleMegh-2296.pdf (filename),usctheses-c3-369633 (legacy record id)
Legacy Identifier
etd-TuttleMegh-2296.pdf
Dmrecord
369633
Document Type
Dissertation
Format
application/pdf (imt)
Rights
Tuttle, Meghen Miles
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
BrainSuite
conservatory
MRI
neuroanatomy
neuroplasticity
Training
VBM