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
/
Contextual interference in motor skill learning: an investigation of the practice schedule effect using transcranial magnetic stimulation (TMS)
(USC Thesis Other)
Contextual interference in motor skill learning: an investigation of the practice schedule effect using transcranial magnetic stimulation (TMS)
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
CONTEXTUAL INTERFERENCE IN MOTOR SKILL LEARNING: AN
INVESTIGATION OF THE PRACTICE SCHEDULE EFFECT USING
TRANSCRANIAL MAGNETIC STIMULATION (TMS)
by
Chien-Ho Lin
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(BIOKINESIOLOGY)
August 2007
Copyright 2007 Chien-Ho Lin
ii
ACKNOWLEDGEMENTS
Although the title page suggests this document was the work of one person, I
would like to acknowledge and recognize the contributions of many individuals who
helped bring this final dissertation together and who supported my long arduous
journey through my PhD program. First of all, this dissertation is the product of more
than a hundred selfless individuals who cheerfully volunteered their time and effort to
help me obtain the information described herein.
I am indebted to my advisor Dr. Carolee Winstein, for the time and guidance
she has devoted to me. Her inquisitive mind, devotion to details, and drive for
excellence has taught me the characteristics needed to become a great scientist. She
will be an everlasting role model for me. I have enjoyed being her advisee and always
learned tremendously from every single meeting with her. I am also grateful for the
contributions of my dissertation committee members: Drs. Stanley Azen, James
Gordon, Kathy Sullivan, and Allan Wu. This work is richer and stronger because of
their brilliant suggestions and contributions. Dr. Azen has been tremendously helpful
for the data analysis piece of this work. I give special thanks to Dr. Gordon, for the
time he has devoted to me for the research design, data processing, and manuscript
writings. I often went to his office with headaches but left with relief and solutions.
I am thankful for Dr. Allan Wu, who is impressively smart but yet is always
open-minded and really listens to others. Working with him as a research assistant
and doctoral student has taught me to be modest. I also want to thank Dr. Sullivan’s
for her guidance and support throughout my PhD study. She is not just a great
iii
guidance committee member but also a great senior. I still vividly remember that she
invited me over to her house on a beautiful summer afternoon to discuss research ideas,
at the time I was getting ready for the doctoral qualifying exam.
I am thankful for the support and friendship of my colleagues in the Motor
Behavior and Neurorehabilitation Laboratory—for those who have gone before—Drs.
Pan Onla-or, Dorian Rose, and Jarugool Tretriluxana and for those who are following
along this path—Jill Stewart, Maureen Whitford, Shailesh Kantak, Caroline Tan, Shu-
ya Chen, Erica Pistch, Hui-ting Goh, and Yi-An Ko. I give special thanks to Jool
(Jarugool Tretriluxana) for always being there for me. She is a great senior classmate
and a great role model. My life as a PhD student has become so much easier because I
knew nothing would go wrong if I just followed her steps. I have been fortunate
beyond measure to observe and benefit from the minds and hearts of a dear mentor
and friend, Dr. Beth Fisher, both in academic and personal perspectives. She has
taught me that life of a great scientist can be so interesting and colorful. I cannot
imagine the five-year study here without her. I have learned so much and look
forward to future work with her.
The energy, support, and laughter of my labmates, Chelle Prettyman,
Samantha Underwood, Patricia Pate, Cindy Kushi, and Chris Hahn have made my
endeavors throughout my studies much more fulfilling and enjoyable. I thank Cindy
Kushi for her cheerfulness and kindness. She is like my “American mom” who is
always so sweet and encouraging to me. I am also grateful to have Chris Hahn around
iv
who is always there whenever I need help and suggestions about participant
coordination and lab logistics.
I would like to give special thanks to Dr. Nina Bradley who was the instructor
of a course entitled, “Advanced Neuroscience,” which I took when I was a first year
PhD student. The first year, especially this required course, had not been easy for me
and quite often I had doubts if I was able to get through it. Dr. Bradley wrote me a
personal email after the midterm to encourage me and tell me how well I had done in
this particular exam. I guess up to now….she has no idea how important that email
was to me. Her email gave me the confidence and courage to know that as long as I
kept working hard and never gave up, nothing was impossible.
I also want to thank David Lord, Matt Sandusky, and Chad Louie who have
helped me solve all the problems that happened in the process of data collection for
this dissertation work. I have special thanks to Matt, who always works with a crazy
schedule and is like a ‘911’ for me. Whenever I was in trouble and needed technical
support, he was always there for me, no matter how late I called.
I am grateful to my family and colleagues in Taiwan, whose love, support, and
encouragement kept me grounded, balanced and moving forward. I owe a debt of
gratitude to my brother, who stayed in Taiwan to accompany my parents so that I
could came overseas to pursue my PhD. Finally, I reserve my deepest gratitude and
respect for my parents for their unfailing and unconditional love and faithful support
through this endeavor from beginning to end. They stood by me and never stopped
believing in me. We kept close communication and shared through daily phone calls
v
over the past five years. Their encouragement has made me strong and persistent.
Fellow students in the lab often ask me why I always appear to be carefree…and I
guess my family is the answer. In these years, I have realized how fortunate I am to
have such a supportive family.
Funding for this work was provided by the Division of Biokinesiology and
Physical Therapy at the University of Southern California, the North American
Society for the Psychology of Sport and Physical Activity, and the California Physical
Therapy Association.
vi
TABLE OF CONTENTS
ACKNOWLEDGEMENTS ii
LIST OF TABLES vii
LIST OF FIGURES viii
ABSTRACT x
CHAPTER 1: BACKGROUND AND SIGNIFICANCE 1
CHAPTER 2: TASK PRACTICE ORDER AND THE ROLE OF HUMAN 8
MOTOR CORTEX IN MOTOR SKILL LEARNING: AN
INVESTIGATION USING TRANSCRANIAL MAGNETIC
STIMULATION (TMS)
CHAPTER 3: THE ROLE OF HUMAN MOTOR CORTEX IN LEARNING 30
MOVEMENT KINEMATICS: AN INVESTIGATION USING
TRANSCRANIAL MAGNETIC STIMULATION (TMS) AND
TASK PRACTICE ORDER
CHAPTER 4: THE CONTEXTUAL INTERFERENCE EFFECT: 46
ELABORATION- DISTINCTIVENESS OR FORGETTING-
RECONSTRUCTION? A POST-HOC ANALYSIS OF TMS-
INDUCED EFFECTS ON MOTOR LEARNING
CHAPTER 5: SUMMARY AND GENERAL DISCUSSION 62
REFERENCES 76
vii
Table 1
Table 2
LIST OF TABLES
Baseline demographic information
Baseline demographic information
12
37
viii
Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
LIST OF FIGURES
Experimental set-up with lever arm, feedback display, and three arm
movement trajectories with specific time and amplitude requirements
Timeline of each event within a single trial in millisecond (ms). For
the TMS and Sham-TMS groups, single TMS pulse was timed to the
onset of inter-trial interval. FB: post-trial feedback
Mean root mean square error (RMSE) of the six experimental groups
during acquisition phase. Blocked: blocked-order training, Random:
random-order training
Average root mean square error (RMSE) of the delayed recall phase.
The X axis is the Practice order. * indicates the group differences are
statistically significant. The RMSE of the recall phase are collapsed
across the two testing conditions. Notice that the benefit of random
practice over blocked practice in motor learning was demonstrated in
the no-TMS control condition (open circles). There was a significant
difference between the Random-control and Blocked-control groups.
Accuracy in movement timing was quantified using absolute error in
movement time (AE_MT). Accuracy in movement amplitude was
quantified using the sum of absolute error in movement amplitude
derived from the three reversal points within each movement trajectory
(AE_SumAmp). AE-ampPeak1: absolute value of amplitude error at
the peak1; AE-ampTrough: absolute value of amplitude error at the
trough; AE-ampPeak2: absolute value of amplitude error at the peak2.
Timing error (Y axis) of the six experimental groups across
acquisition and retention phases. The acquisition includes Block 1 to
12; Block 13 is the second-day retention phase. The timing error of
retention phase was the average error of two testing conditions (i.e.
blocked and random). * indicates the differences are statistically
significant.
Amplitude error in degree of the six experimental groups across
acquisition and two retention phases. The acquisition includes Block
1 to 12; Blocks 13 is the immediate retention phase; Blocks 14 is the
delayed retention phase. The RMSE of retention tests are the average
after clasping two testing conditions (i.e. blocked and random)
14
14
18
20
36
39
40
ix
Fig. 8
Fig. 9
Fig. 10
Fig. 11
Predicted results of the experimental manipulation based on
Elaborative Processing Hypothesis (A) and Forgetting
Reconstruction hypothesis (B). TMS: disruptive transcranial
magnetic stimulation pulses
Mean root mean square error (RMSE) of the four experimental
groups during acquisition (block 1 to 12), immediate (block 13),
and delayed retention (block 14) phases. Blocked: blocked
practice, Random: random practice. * indicates the group differences
are statistically significant. The RMSE of the retention phase are the
presented after collapsing two testing conditions.
(A) Mean root mean square error (RMSE) of the acquisition phase.
The X axis is the Practice order. (B) RMSE of the delayed retention
phase. * indicates that the group differences are statistically
significant.
Mean decrement in performance from day1 to day2, defined as the
mean RMSE changes between the immediate and delayed retention
phases. The X axis is the Practice order. * indicates that the group
differences are statistically significant.
49
54
55
57
x
ABSTRACT
This dissertation was designed to investigate the neural basis implementing the
contextual interference effect in motor skill learning. Sixty-one non-disabled adults
were recruited. Participants practiced three fast, discrete, goal-directed arm
movements each with specific time and amplitude requirements. The motor tasks
were practiced either in a blocked or quasi-random order. Transcranial magnetic
stimulation (TMS) was applied to the arm areas of the cortical motor system (CM) to
directly perturb brain processing during motor practice. Single TMS pulses were
delivered, synchronized to each inter-trial interval. The three stimulation conditions
(no TMS Control, TMS, Sham) and two practice orders (Blocked, Random) factorial
design resulted in six experimental groups. Testing took place over 2 consecutive
days with acquisition and immediate retention phases on day 1 and a delayed retention
phase on day 2. The retention tests consisted of trials without feedback with which
neither TMS nor Sham-TMS applied.
This dissertation is organized around a single experiment designed to answer
three separate, yet related research questions (Chapters 2, 3, and 4). Each will be
presented separately with its own unique purpose, methods, results and discussion.
Overall, results from this dissertation provide relevant information about the
neuromotor and psychological explanations for the CI effect in motor learning. We
demonstrated that cortical processing, specifically that involving the arm area of the
cortical motor system, is a putative neural locus for the CI effect. In addition to
supporting previous findings that human motor cortex contributes to motor skill
xi
learning, these results suggest that the level of CM engagement can be modulated by
task practice order. With this experimental design, we also tested the two strongest
information processing explanations for the CI effect. These findings have important
implications for understanding the nature by which fast discrete goal-directed arm
movements are best learned. Our work also provides new insights into a beginning
foundation for the translation of motor learning principles into advanced and
developing therapeutic approaches in neurorehabilitation.
1
CHAPTER 1
BACKGROUND AND SIGNIFICANCE
The Phenomenon
The contextual interference (CI) effect, as defined by Hall and Magill (1990),
is the effect on learning of functional interference in a practice situation when several
tasks must be learned and are practiced together (Magill and Hall, 1990, Hall and
Magill, 1995). During practice, high levels of interference occur such as when
multiple tasks are practiced in a quasi-random order (i.e., C-A-B-, A-B-C-, B-C-A-, if
each letter represents a practice task; to simplify the description, I will use “random order”
to represent “quasi-random order” from now on), usually resulting in less effective
performance during acquisition than low interference conditions, in which all or a set
of trials on one task are practiced before participants switch to the next task (i.e., A-A-
A-, B-B-B-, C-C-C-). However, when performance is assessed during delayed
retention or transfer tests, the participants who practiced under a random order
condition demonstrate superior performance compared to those who practiced in a
blocked order condition. This phenomenon is well supported in verbal learning
(Battig, 1979) but was not tested with motor skill learning until Shea and Morgan’s
now classic study published in 1979 (Shea and Morgan, 1979).
Since the Shea and Morgan study, the learning benefit of a random practice
order, as compared with a blocked practice order has proven to be a repeatable and
robust phenomenon, at least for the learning of simple motor skills in laboratory
2
situations using tasks including: 1) multiple-segments with participants required to
move as fast as possible (Lee and Magill, 1983, Shea and Zimny, 1983, Gabriele et al.,
1987, Limons and Shea, 1988); 2) multiple-segments with participants required to
meet certain overall movement time criteria (Gabriele et al., 1987, Carnahan et al.,
1990); 3) multiple-segments with participants required to produce certain segment
times (Lee and Magill, 1983, Lee and Magill, 1985, Lee et al., 1985, Lee et al., 1992,
Wulf and Lee, 1993, Hall and Magill, 1995, Sekiya et al., 1996, Lee et al., 1997, Shea
et al., 2001, Giuffrida et al., 2002); 4) simple aiming (Young et al., 1993); 5)
anticipation timing (Del Rey, 1982, Del Rey et al., 1982, Del Rey, 1989, Del Rey et al.,
1994); 6) movement patterning (Wulf, 1992, Sekiya et al., 1996); 7) tracking (Jelsma
and Pieters, 1989, Jelsma and Van Merrienboer, 1989); 8) dart throwing (Meira and
Tani, 2001); 9) rapid force production (Shea et al., 1990); 10) pattern drawing (Albaret
and Thon, 1998); 11) spatial error-detection (Sherwood, 1996); and 12) computer
games (Shewokis, 1997).
When the CI effect was tested using more complex, “real-life” skills, the
benefit of random practice was evident in several activities, such as badminton serves
(Wrisberg and Liu, 1991, Goode et al., 1998), kayaking (Smith and Davies, 1995),
rifle shooting (Boyce and Del Rey, 1990), volleyball skills (Bortoli et al., 1992),
baseball batting (Hall et al., 1994), and bimanual limb movements (Tsutsui et al.,
1998). In summary, these studies provide convincing evidence for the greater
effectiveness for motor learning of randomly ordered practice compared with blocked
ordered practice. These studies also highlight the well-known learning-performance
3
distinction, and emphasize that not all immediate changes during the practice phase
can be assumed to reflect the relatively permanent changes in behavior, defined as
learning. The learning-performance distinction is analogous to what one might see in
the therapeutic clinic setting when a patient is able to perform a motor task quite well
but that performance is seen to deteriorate once he/she goes home; further, a patient
may not be able to generalize or transfer the performance from the clinic to functional
daily activities in the home environment. Consequently, performance during the
acquisition phase may not be a very good representation of what was actually learned;
motor learning is best assessed by performance during a delayed recall test at a time
somewhat removed from the acquisition phase (i.e., retention phase).
The benefit of random order practice for motor learning has been explained in
part by the more elaborate task-relevant information processing that is evoked, such as
inter-task comparisons and action planning, that induces more comprehensive and
retrievable memory traces. This type of information processing is thought to be
particularly beneficial for supporting subsequent retrieval efforts that occur during
delayed recall (Lee and Magill, 1983, Shea and Zimny, 1983, Lee and Magill, 1985,
Hall and Magill, 1995). Two principle explanations for the CI effect are the (1)
elaborative processing and (2) forgetting-reconstruction hypotheses. The elaborative
processing hypothesis holds that random practice induces the learner into more
elaborative processing such as inter-task comparisons and embellishment of task-
relevant information (Shea and Morgan, 1979, Shea and Zimny, 1983, Shea and
Zimny, 1988, Shea and Titzer, 1993). More elaborative information processing is
4
thought to result in more comprehensive and readily retrievable memory traces.
According to the elaborative processing view, the interval prior to the next movement
trial (i.e., inter-trial interval) is a time during which inter-task comparisons are made
and some amount of elaboration likely occurs.
Alternatively but not mutually exclusive, the forgetting reconstruction
hypothesis suggests that a previously constructed action plan is more likely to be
available in working memory during blocked order practice, since the same task is
practiced repeatedly (Lee and Magill, 1983, Lee and Magill, 1985). Random practice
however, induces the learner to abandon the previously constructed action plan
because he or she has to perform a somewhat different task on the next trial. The
forgetting that ensues requires the learner to actively reconstruct the action plan the
next time the initial task is practiced; this results in deeper processing and a stronger
memory representation. As with the elaborative processing hypothesis, forgetting and
reconstruction of the action plan are likely to occur in the interval between practice
trials.
Though previous studies have focused on the information processing during
task practice such as encoding and retrieval, little work has focused on the neuromotor
mechanism which implements the CI effect. Therefore, the overall purpose of my
dissertation was to probe the neural mechanism of the CI effect through application of
transcranial magnetic stimulation (TMS) to directly perturb the neural processes
evoked during the inter-trial interval. TMS is a technique that can enhance or inhibit
cortical reorganization (Ziemann et al., 1998) and information processing (Flitman et
5
al., 1998b, Boroojerdi et al., 2001) by means of noninvasive focal magnetic
stimulation of the human brain (Hallett, 2000). Consequently, one strategy for
probing the neural mechanism of the CI phenomenon in motor learning is to disrupt
essential information processing with focal TMS pulses delivered to the primary motor
area of the cerebral cortex. Such a paradigm is similar to in vitro experiments in which
stimulation of cortical afferents are paired with a depolarization of the synaptic target
neuron under specific temporal conditions that mimic an LTP- or LTD-like
stimulation (Baranyi and Szente, 1987, Baranyi et al., 1991, Butefisch et al., 2004). In
the present study, we tested the role of the cortical motor system in learning goal-
directed arm movements under either a blocked or random order practice condition.
The cortical motor system (CM), which includes the premotor (PMd), primary motor
(M1), and somatosensory (S1) cortices, was chosen primarily for consideration of
TMS specificity. A single-pulse TMS perturbation was delivered precisely during
each inter-trial practice interval to the motor representation of the primary muscle
required to perform the movements. By synchronizing the TMS pulse during this
interval, we expected to reveal CM’s putative role in the hypothesized processes as
they relate to motor learning.
This dissertation has four distinct yet interrelated specific aims:
Aim 1: To determine if the CI effect can be generalized to the learning of fast
discrete goal-directed arm movements each with specific spatial and
temporal requirements.
6
Aim 2: To investigate the role of the cortical motor system (CM) in the CI
effect on motor learning through direct perturbation of CM processes
during motor practice using transcranial magnetic stimulation (TMS).
Aim 3: To investigate the role of the cortical motor system (CM) through
kinematic analyses of task-relevant parameters which are
tuned/specified during skill acquisition.
Aim 4: To determine which of the two information processing explanations
(Elaborative-Processing and Forgetting-Reconstruction Hypotheses)
best accounts for the CI effect in motor learning.
Overview
This dissertation is organized around a single experiment to answer three
separate, yet related research questions; each will be presented separately with its own
unique purpose, methods, results and discussion. All three questions examine the CI
effect in motor learning with non-disabled adults whose age ranged from 18 to 37
years old. Research Question 1 investigated if the CI effect can be generalized to the
learning of a set of fast discrete goal-directed arm movements with specific spatial and
temporal requirements (Chapter 2). Next, transcranial magnetic stimulation (TMS)
was applied to study the role of the cortical motor system (CM) in the CI effect
(Chapter 2). The results described in Chapter 2 supported the hypothesis that CM has
a direct role in implementing the CI effect. This allowed the retrospective kinematic
analyses specific to movement timing and amplitude to be applied to examine the
nature of CM’s role (Chapter 3). The research question addressed in Chapter 3 is how
task-relevant parameters are tuned and/or specified in the process of motor learning.
7
In addition, the experimental design allowed me to test the unique predictions
of each of two information processing explanations (Elaborative-Processing and
Forgetting-Reconstruction Hypotheses) for the CI effect (Chapter 4). The
experimental manipulation using a TMS perturbation should lead to one of two
different results, depending on whether elaborative processing or forgetting-
reconstruction provides the better explanation for the contextual interference effect. If
elaborative processing accounts for the effect, then the advantage of random practice
should be diminished by the TMS, since the TMS would interfere with elaborative
processing during the inter-trial interval. Alternatively, if forgetting and reconstruction
accounts for the contextual interference effect, blocked practice with TMS should
become more like random practice, since the TMS applied during the inter-trial
interval could induce the learner to reconstruct the action plan even during blocked
practice in which the same task is practiced repeatedly.
Chapter 2, Chapter 3, and Chapter 4 at the time this dissertation was written
are either under review or in preparation for submission to selected interdisciplinary
journals. The three manuscripts are reprinted here for completeness. Chapter 5 is
constructed as a general discussion and summary of the dissertation with its
limitations, implications and future directions.
8
CHAPTER 2
TASK PRACTICE ORDER AND THE ROLE OF HUMAN MOTOR CORTEX
IN MOTOR SKILL LEARNING: AN INVESTIGATION USING
TRANSCRANIAL MAGNETIC STIMULATION (TMS)
Introduction
Previous literature has shown that motor learning is enhanced when practice
variability within session is higher than that achieved in constant practice conditions.
One way to manipulate practice variability while keeping the number of trials and
tasks constant is to change the order in which tasks are practiced. For example, a
random practice order in which tasks are practiced in a quasi-random order (i.e., C-A-
B-, A-B-C-, B-C-A-, each letter represents a task) can be compared to a blocked
practice order in which each task is practiced repeatedly prior to switching to the next
task (i.e., A-A-A-, B-B-B-, C-C-C-). Results from studies investigating the effects of
practice order on motor learning typically show that a random practice order enhances
motor learning (i.e., performance on delayed recall tests is more accurate) compared
with a blocked practice order (Shea and Morgan, 1979, Hall and Magill, 1995,
Goodwin and Meeuwsen, 1996, Lee et al., 1997, Meira and Tani, 2001, Giuffrida et al.,
2002, Perez et al., 2005, Wright et al., 2005). This benefit has been explained in part
by the more elaborate information processes that are evoked through such operations
as inter-task comparisons and generative processes including action planning. These
more elaborative information processes evoked by random practice are thought to
9
result in more comprehensive and retrievable memory traces that are particularly
beneficial for supporting subsequent recall efforts (Lee and Magill, 1983, Shea and
Zimny, 1983, Lee and Magill, 1985, Lee, 1988, Hall and Magill, 1995). What remains
to be determined, however, is the brain-behavior relationship in this context-specific
motor learning. This study directly examines the interaction between the human
cortical motor system and task order variability encountered during motor practice.
The role of the cortical motor system in motor learning as conceptualized in
the Control-Based Learning Theory (Willingham, 1998) suggests that motor learning
grows directly out of motor control processes. This theory predicts, that in addition to
its direct roles in motor planning and execution, the cortical motor system, which
includes the premotor (PMd), primary motor (M1), and somatosensory (S1) cortices,
also play direct roles in motor learning. Motor learning emerges as the cortical motor
system becomes more efficient in kinematics specification, visuomotor transformation,
motor program selection, and feedback processing (Tanji, 1996, Zhang et al., 1997,
Hund-Georgiadis and von Cramon, 1999, Lee et al., 1999, Meister et al., 2005,
Dancause, 2006, Floyer-Lea et al., 2006, Lee and van Donkelaar, 2006, Praeg et al.,
2006).
Likewise, neurophysiologic and neuroimaging studies have provided
substantial evidence that the cortical motor system (CM) is significantly engaged
╫
during motor skill learning, particularly when the demands during motor practice are
challenging or variable (Mitz et al., 1991, Elbert et al., 1995, Nudo et al., 1996, Cohen
et al., 1998, Karni et al., 1998, Classen et al., 1999, Hess et al., 1999, Hund-Georgiadis
10
and von Cramon, 1999, Liepert et al., 1999, Frost et al., 2000, Plautz et al., 2000,
Kleim et al., 2003, Ben-Shaul et al., 2004, Butefisch et al., 2004, Hatfield et al., 2004,
Pleger et al., 2004, van Mier et al., 2004, Ziemann et al., 2004, Kleim et al., 2006).
Reorganization of movement representations occurs when tasks are sufficiently
challenging as to require skilled action for successful accomplishment (e.g., successful
retrieval of food pellets from a small well or a rotating disk) (Nudo et al., 1996, Kleim
et al., 1998, Plautz et al., 2000, Kleim et al., 2002). Similarly, in human imaging
studies using functional magnetic resonance imaging (fMRI), positron emission
tomography (PET), or transcranial magnetic stimulation (TMS), greater engagement
of CM is observed when subjects attempt challenging motor tasks or are trained in a
variable practice condition (Demiralp et al., 1990, Zhang et al., 1997, Sergio and
Kalaska, 1998, Chung et al., 2000, Schubotz and von Cramon, 2002, Gorbet et al.,
2004, Hatfield et al., 2004, Lewis et al., 2004, Seidler et al., 2004, Aoki et al., 2005,
Meister et al., 2005). These studies provide evidence that variability during motor
practice elicits more engagement of CM than does repetitive, consistent motor practice.
Such experience-dependent CM activity in motor learning suggests that in addition to
its known role in movement execution, CM activity is also associated with motor skill
learning, particularly when practice variability increases.
Given that CM activity associated with motor learning likely depends on the
degree of variability during motor practice, we reasoned that CM would engage to a
greater extent in a random order condition (more variable) than a blocked order
condition (less variable). We tested the causal significance of an increased
11
contralateral CM activity during motor practice on motor learning using transcranial
magnetic stimulation (TMS). TMS is a technique that can introduce a transient
perturbation to cortical information processing by means of noninvasive focal
stimulation of the human brain (Hallett, 2000). Consequently, one strategy to probe
the interaction between CM and practice variability is to disrupt CM processes with a
TMS perturbation paradigm. A single disruptive TMS pulse was delivered during
each inter-trial interval during motor practice. The provision of feedback from a
previous trial is thought to inform the elaborative comparison between trials. These
feedback-related processes are thought to operate during the interval just after
feedback is presented (i.e., the inter-trial interval) (Lee and Magill, 1983, Shea and
Zimny, 1983, Lee and Magill, 1985, Hall and Magill, 1995, Lee et al., 1997). By
disrupting CM processes associated with the primary agonists for the motor tasks
during this critical interval, we expected to perturb CM function associated with the
putative processes for motor learning.
We hypothesized that disruptive TMS pulses applied to CM during motor
practice would influence the learning of the practiced arm tasks. To clarify the effect
of TMS on motor learning (not simply a transient effect on performance), we assessed
the learning effect using a delayed recall test administered one day after practice.
Since greater information processes are thought to take place during the inter-trial
interval under random compared with blocked practice (Battig, 1979, Lee and Magill,
1983, Shea and Zimny, 1983, Lee and Magill, 1985, Gabriele et al., 1987, Limons and
Shea, 1988, Wright et al., 1992, Lee et al., 1997, Wright et al., 1997, Wright et al.,
12
2005), the strategically timed disruptive TMS pulse was expected to be more
detrimental to motor learning under random compared with blocked practice. To
account for any non-specific effects during the TMS procedure (e.g., TMS coil noise),
we also included Sham-TMS conditions in our design.
Materials and Methods
Subjects
Sixty one healthy volunteers (18 - 38 years) who were naive to the task were
recruited (Table 1). All subjects gave written informed consent and participated in the
study under a protocol approved by the Institutional Review Board of the University
of Southern California Health Sciences Campus.
Table 1. Baseline demographic information of Experiment 1
Stimulation
Condition
Control (no TMS) Sham-TMS TMS
Training schedule Blocked Random Blocked Random Blocked Random
Age 26.5±1.728.4±1.424.5±1.124.6±1.126.4±1.3 28.4±2.1
Gender 4M6F 6M4F 7M3F 5M6F 6M4F 3M7F
Baseline MT - - - - 61.9±2.6 59.7±3.4
MT: resting motor threshold, is given as a percentage of maximal stimulator output.
MT is the minimal stimulating intensity to elicit an isolated movement of the biceps
muscle.
Values are means ± SD.
Exclusion criteria included all contraindications to receiving transcranial
magnetic stimulation (Wassermann, 1998), specifically the presence of a pacemaker,
metal in the head, pregnancy, any neurological disorder, current use of stimulants or
medications known to lower seizure threshold, and personal or family history of a
seizure disorder.
13
Instrumentation and Task
A lightweight lever affixed to a frictionless vertical axle was attached to a table
and positioned parallel to the floor. A handle at the end of the lever was adjusted to
accommodate the length of participant’s forearm. A linear potentiometer was attached
to a transducer at the base of the vertical axle. Signals from the transducer were
converted to a digital signal by an A-D board of a Compaq 466v computer and
sampled at 1000 Hz. A LabView-based (Labview, National Instruments) custom-
made software program (Weekley, 2004) was applied to manipulate the movement
trajectory, the timeline of each event, and data storage for each trial. The motor task
was to move the lever at the correct speed and distance to replicate a goal movement
trajectory that was displayed on the computer monitor before each trial (Fig. 1). The
goal was to learn three discrete movement trajectories where each movement was
composed of two arm extension-flexion reversal actions (Fig. 1).
During the acquisition phase, the participants were presented with feedback
after each movement trial. The feedback included 1) an overall error score, root mean
square error (RMSE), representing the difference between the goal movement
trajectory and the participants response and 2) a graphic representation of the
participant’s response superimposed on the goal movement trajectory. All participants
received written instructions and additional verbal information about the experiment.
The timeline of each event during a single trial is illustrated in Figure 2. The goal was
to learn three discrete movement trajectories where each movement was composed of
two arm extension-flexion reversal actions.
14
Three Task Trajectories
Figure 1. Experimental set-up with lever arm, feedback display, and three arm
movement trajectories with specific time and amplitude requirements.
Figure 2. Timeline of each event within a single trial in millisecond (ms). For the
TMS and Sham-TMS groups, single TMS pulse was timed to the onset of
inter-trial interval. FB: post-trial feedback.
Action FB Delay FB presentation
5000 2000
5000
Inter-trial interval
Ready Go
1000
Target
Display
200
Trial 1 Trial 2
Target
Display
TMS
2000
2000
Monitor
Top-down view
15
Experimental Design:
Participants were randomly assigned to one of six practice conditions: no TMS
control (Blocked-Control, Random-Control), TMS (Blocked-TMS, Random-TMS),
and Sham-TMS (Blocked-Sham TMS , Random-Sham TMS). TMS pulses were
applied to the motor cortex representation contralateral to agonist muscles used to
perform the tasks. The Sham-TMS condition (i.e., coil applied to same location,
normal clicking noise apparent, but no magnetic pulse applied) was implemented to
control for the non-specific effects of TMS including coil noise and scalp sensation.
The three arm tasks were practiced in either a blocked or random order for 144 total
trials. During blocked practice, there were 48 movement trials of each arm task and
the order of tasks were counter-balanced across participants. During random practice,
the three tasks were practiced in a non-repetitive manner within each 48-trial set; the
randomization scheme was the same for all random practice subjects.
Experimental Procedure
Since the learning performance distinction is a well-recognized phenomenon
in learning and memory research, we employed a recall-test paradigm in order to
measure the relatively permanent effects on motor performance (Shea and Morgan,
1979, Hall and Magill, 1995, Goodwin and Meeuwsen, 1996, Lee et al., 1997, Cahill
et al., 2001, Meira and Tani, 2001, Giuffrida et al., 2002, Perez et al., 2005, Schmidt
and Lee, 2005, Wright et al., 2005). Testing was conducted over two consecutive days
with acquisition and immediate recall phases on day 1 and a delayed recall phase on
day 2. To ensure that the recall test was not biased toward either practice order,
16
participants were given 12 recall trials each under both blocked and random order
conditions and the recall test order was counterbalanced across participants. All
participants completed an immediate and delayed recall phase without post-response
feedback or the application of TMS, real or sham.
Magnetic Stimulation
Magnetic pulses were delivered by a magnetoelectric stimulator (MES-10,
Cadwell Laboratories, Kennewick, WK) through a circular magnetic coil (9.0 cm
diameter). In considering the duration of the practice period (144 trials which lasted
for around 90 minutes), a circular coil was chosen rather than a figure-8 as there were
less overheating problems with this coil. The coil was placed tangentially to the scalp
in a direction such that monophasic pulses induced a current flowing antero-
posteriorly perpendicular to the central sulcus (Werhahn et al., 1994, Kaneko et al.,
1996). To center our stimulation over the motor cortex, we placed one edge of the coil
on the subject’s scalp and identified an optimal location where stimuli consistently
elicited a maximal motor evoked potential (MEP) in the contralateral biceps brachii
muscle (Kamen, 2004). Motor threshold was defined as the minimum TMS intensity
to the nearest 1% of maximum stimulator output that evoked a MEP ≥ 50 μV in at
least 5 out of 10 trials in the target muscle (Rossini et al., 1994). TMS pulses were
only applied during the acquisition, practice phase and synchronized to the onset of
each inter-trial interval (Fig. 2). This interval was chosen as a critical period for
cognitive processes during practice when information from the previous trial might be
used in the preparation for the next trial (Pascual-Leone et al., 1992b, Berardelli et al.,
17
1994, Taylor et al., 1995, Ziemann et al., 2004). A suprathreshold TMS intensity
(120% of resting motor threshold) was chosen in order to produce a disruptive
influence on cortical activity centered on and around the motor cortex.
Dependent Measures
Overall performance accuracy was quantified using root mean square error
(RMSE) (Wulf et al., 1993). RMSE is the average difference between the goal
trajectory and the participant’s response calculated over the participant’s total
movement time. Online RMSE was calculated after synchronizing the onset of the
target trajectory with the onset of the subject’s movement.
Offline, an average RMSE was computed for every twelve-trial block of
acquisition performance to assess improvements during practice. An average RMSE
for immediate recall was computed from performance 10 minutes following practice
to assess the short-term recall ability. An average RMSE for delayed recall was
computed from day 2 performance as a reliable measure of the relatively long-term
recall ability.
Statistical Analysis
For the acquisition data, average RMSE was submitted to a 3 Stimulation
Condition (no TMS Control, TMS, Sham-TMS) × 2 Practice order (Blocked, Random)
× 12 Practice Block analysis of variance (ANOVA) with repeated measures for
practice block. For the retention data, a 3 (Stimulation Condition) × 2 (Practice order)
ANOVA was conducted separately for each recall phase. When appropriate, and for
determining the locus of any significant interaction, the ANOVAs were followed by a
18
simple main effect test and Post hoc Tukey test to control for inflated Type 1 error.
For all statistical tests, significance level was set at p < .05. SPSS 13.0 (SPSS Inc.,
Chicago, IL) statistical software was used for all statistical analysis.
Results
Group mean comparisons for age, gender and resting motor threshold are
summarized in Table 1. There were no significant group differences for age or gender.
The resting motor threshold prior to motor practice did not differ significantly between
Blocked-TMS and Random-TMS groups.
Figure 3. Mean root mean square error (RMSE) of the six experimental groups
during acquisition phase. Blocked: blocked-order training, Random:
random-order training. Error bars are standard errors of mean.
19
Acquisition phase
Figure 3 illustrates the group average RMSE for the acquisition phase. At
baseline (block 1), there was no group difference in motor performance, p = .467 (Fig.
3). Each group showed an improvement (decrease in RMSE) over time that resulted
in a significant main effect for Practice Block, F(11, 45) = 27.923, p = .000. There
was a significant Stimulation Condition by Practice order interaction, F(2, 55)= 3.534,
p = .036. TMS and Sham-TMS conditions appear to have enhanced acquisition
performance of blocked practice but deteriorated that of random practice. However,
post hoc tests indicated that this stimulation condition effect was not significant for
either practice-order condition.
Retention Phase
Immediate Recall
Motor practice with TMS or sham TMS did not influence performance during
the immediate recall phase. The effects of Stimulation Condition and the interaction
between Stimulation Condition and Practice order were not significant.
Delayed Recall
Average RMSE during delayed recall showed a significant Stimulation
Condition by Practice order interaction, F(2, 55)= 5.899, p= .027 (Fig. 4). Post-hoc
tests showed that the Random-Control group exhibited better recall performance than
the Blocked-Control group, p = .01 (Fig. 4, open circles). This finding of enhanced
motor learning following random practice replicates what has been shown previously
(Shea and Morgan, 1979, Hall and Magill, 1995, Goodwin and Meeuwsen, 1996, Lee
20
et al., 1997, Meira and Tani, 2001, Giuffrida et al., 2002, Perez et al., 2005, Wright et
al., 2005).
Further and particularly relevant to this study, the mean RMSE for the
Random-TMS group was significantly higher than that for the Random-Control (p
= .01) and the Random-Sham TMS (p = .02) groups (Fig. 4, right). However, this
pattern was not observed for the blocked practice group (Fig. 4, left), suggesting that
TMS had a selective effect on the learning of motor tasks for the random group.
Figure 4. Average root mean square error (RMSE) of the delayed recall phase.
Error bars are standard errors of mean. The X axis is the Practice order.
* indicates the group differences are statistically significant. The
RMSE of the recall phase are collapsed across the two testing
conditions. Notice that the benefit of random practice over blocked
practice in motor learning was demonstrated in the no-TMS control
condition (open circles). There was a significant difference between
the Random-control and Blocked-control groups.
21
Discussion
To our knowledge, this is the first study using TMS as a cortical perturbation
during motor practice to perturb the interaction between practice context and
subsequent motor learning. Given that random practice is thought to evoke greater
information processes than blocked practice, we reasoned that a disruption of cortical
motor activity during practice would have a greater effect on delayed recall
performance for random rather than blocked practice.
Our main finding is that TMS perturbation centered over the contralateral
motor cortex during motor practice interfered with motor learning for subjects who
practiced under the random order condition but not those who practiced under the
blocked order condition. This result suggests that the cortical regions at and around
the motor cortex (corticomotor system, CM) are not only a substrate for motor
learning but also that these neural processes are dependent on the within-session
variability engaged during motor practice (higher variability is inherent in random
practice than in blocked practice). The results also suggest that CM participates in the
consolidation of motor memory for task-relevant information. The findings will be
discussed within the context of CM engagement in blocked and random practice,
followed by how these results provide insight into a fundamental principle of
experience-dependent neural plasticity.
Cortical Motor System Perturbation Deteriorate Motor Learning
Though random practice did not enhance performance during the acquisition
phase, its benefit became apparent when performance was re-evaluated after a 24-hour
22
retention interval without TMS (Fig. 4). This suggests that not all changes during the
practice phase represent relatively permanent changes in motor behavior (Cahill et al.,
2001, Schmidt and Lee, 2005), and exemplifies the learning performance distinction
paramount in the skill learning literature (Shea and Morgan, 1979, Hall and Magill,
1995, Goodwin and Meeuwsen, 1996, Lee et al., 1997, Cahill et al., 2001, Meira and
Tani, 2001, Giuffrida et al., 2002, Perez et al., 2005, Schmidt and Lee, 2005, Wright et
al., 2005). We found that the Random-TMS group performed significantly worse than
did the Random-Sham TMS and the Random-Control groups, suggesting that the
influence of TMS during random practice was mediated primarily through cortical-
specific perturbation. In contrast, recall performance for the Blocked-TMS group was
similar to that of the Blocked-Sham TMS group and while both appeared enhanced
over the Blocked-control group, neither TMS nor sham TMS showed significant
differences from the no-TMS control (Fig. 4, left). This dissociation suggests that
TMS applied to the corticomotor system contralateral to the training arm may have
interfered with essential processes evoked by random but not blocked order practice.
This differential effect from the stimulation condition on practice order is reminiscent
of the increased brain alterations in motor representations seen in both human and
animal brains when complex skills are practiced or enriched environments are
encountered (Demiralp et al., 1990, Zhang et al., 1997, Sergio and Kalaska, 1998,
Chung et al., 2000, Plautz et al., 2000, Kleim et al., 2002, Schubotz and von Cramon,
2002, Bestmann et al., 2004, Gorbet et al., 2004, Lewis et al., 2004, Aoki et al., 2005,
Gregori et al., 2005, Meister et al., 2005).
23
Information Processing during Inter-trial Interval is Essential for Motor Learning
A novel finding in our results is that TMS disruption during random practice,
timed to occur 5 seconds following presentation of visual feedback, produced a
significant effect on motor learning that was evidenced the next day (Fig. 4, right).
This finding provides neurophysiologic evidence for a critical time window around the
inter-trial interval following feedback which is essential in random, but not blocked
practice, for motor learning to occur. As such, our results provide some support to the
hypothesis that the processes associated with random practice, such as task-elaboration,
action planning, and consolidation (Lee and Magill, 1983, Shea and Zimny, 1983,
Shea and Wright, 1991, Shea and Titzer, 1993, Lee et al., 1997), are operative during
this critical post-feedback inter-trial interval. The effects of single-pulse TMS are a
combination of both inhibitory and excitatory neurophysiologic elements and are often
used to probe the excitability of the corticomotor system. In this study, the effects
seen with single-pulse TMS are consistent with a disruptive influence as has been
established in other studies of cortical processing (Pascual-Leone et al., 1992b,
Berardelli et al., 1994, Taylor et al., 1995, Amassian et al., 1998, Desmurget et al.,
1999, Van Donkelaar et al., 2000, Ziemann et al., 2004).
On the other hand, single-pulse TMS can also be facilitatory when the pulse is
appropriately timed and the intensity is below motor threshold. In Bütefisch et al.,
sub-threshold single-pulse TMS timed to coincide with a stereotypic finger movement
showed a facilitation of motor learning based on the theory of Hebbian Learning
(Hebb, 1949, Butefisch et al., 2004). Interestingly, an asynchronous application of
24
supra-threshold TMS in our study produced an impairment of this motor learning. In
light of these studies, our stereotypic TMS pulse was unlikely to so precisely coincide
with reinforceable neuronal activities of each trial. As such, any asynchrony with
cortical motor activity by the supra-threshold TMS pulse, as likely was present, would
also be predicted to impair the benefits of random practice.
The exact physiological mechanisms underlying these condition-specific
effects remain to be determined. Regardless, it is notable that single-pulse modulation
applied in the inter-trial interval during the acquisition phase under random-order
practice conditions produced significant effects on motor learning that were revealed
by performance the next day. This conclusion is limited because we did not
investigate the effect of TMS pulses during other time windows. However, producing
a dissociation in cortical motor activity between blocked- and random-order practice
conditions in this particular time window does strengthen the validity of our design.
Future studies that examine the effects from different time windows or with other
neurophysiologic methods such as magnetoencephalography (MEG) or
electroencephalography (EEG) may be helpful to further understand critical time
windows associated with motor learning.
The role of CM in motor learning as conceptualized in the Control-Based
Learning Theory (Willingham, 1998) suggests that motor learning grows directly out
of motor control processes. Our results support the predictions of the Control-Based
Learning Theory by demonstrating that CM has an essential role in motor skill
learning, via processing of task-relevant information such as kinematics specification,
25
visuomotor transformation, and feedback processes (Lee and Magill, 1983, Lee, 1988,
Hall and Magill, 1995, Tanji, 1996, Zhang et al., 1997, Hund-Georgiadis and von
Cramon, 1999, Lee et al., 1999, Meister et al., 2005, Dancause, 2006, Floyer-Lea et al.,
2006, Lee and van Donkelaar, 2006, Praeg et al., 2006).
Cortical Motor System Is Associated with Motor Memory Consolidation
Our results also suggest that the cortical motor system is a neural substrate for
motor memory consolidation (Robertson, 2004, Robertson et al., 2005). We found
that the effect of TMS for the random order practice was not exhibited until the
delayed recall phase. This suggests that the TMS perturbation interfered with
processing of task-relevant information that supports subsequent motor memory
formation. An alternative explanation offers that essential processes are modulated by
CM and somehow related to motor memory storage, but are independent of the
practice condition. This is, however, less likely given that learning in the Blocked-
TMS group was not affected by cortical perturbation to CM. If CM was a neural
substrate for global motor memory storage, one would expect to see similar TMS-
induced disruption for both blocked and random practice conditions. The fact that
TMS did not disrupt the learning under blocked-order practice also suggests that CM
is not the permanent storage place for motor memory.
These explanations do not rule out the possibility that the function of CM
could be perturbed in ways other than cortical stimulation. For example, the
peripheral muscle twitches induced by supra-threshold TMS could have potentially
"perturbed" the cortical motor system through somatosensory input. It is also hard to
26
separate the role of skill practice from that of attentional load in eliciting practice-
induced plasticity. It is likely, however, that all these factors contribute to this
disruptive effect.
Though we placed the TMS coil such that the magnetic field was centered over
the arm representation within the primary motor cortex (M1), the nonfocal nature of
the circular coil produces a current flow that is likely to affect adjacent dorsal
premotor cortex (dPMC) and somatosensory cortex (S1). The dPMC has been
implicated in motor learning through visuomotor associations and through selecting
from a group of motor programs (Lee and van Donkelaar, 2006, Praeg et al., 2006),
factors that are relevant to our study. S1 shares significant overlap in the corticomotor
system with M1 and likely processes somatosensory feedback from the limb with each
movement (Dancause, 2006, Floyer-Lea et al., 2006). While our use of a circular-
shaped coil limits specificity of the cortical regions perturbed, it is clear that our study
design highlights a cortical-specific influence in primary motor or motor planning
cortices that is selectively present during random, but not blocked-order, practice of
motor skills.
Nature of the Motor Tasks May Influence Motor Memory Consolidation
Recently, Muellbacher et al. (Muellbacher et al., 2002) and Baraduc et al.
(Baraduc et al., 2004), using repetitive TMS (rTMS) over the primary motor cortex,
were able to disrupt early memory consolidation of ballistic finger movements, whose
practice was arranged in a blocked order. These results are in contrast with those
shown in our study that learning a coordinated arm movement with a blocked practice
27
order was not disrupted by single-pulse TMS. There are three main reasons that could
give rise to the difference in the effects of TMS disruption on the consolidation of
these two motor tasks. First, the memory of the motor training in the ballistic and
coordinated tasks could be stored in CM, but in a different manner for each task. We
speculate that the significant TMS effect observed in the ballistic tasks might be due to
the increased recruitment of specific corticospinal projections, for example through an
increase in excitability of all neurons that have no inhibitory inputs to the target
muscles as well as no excitatory inputs to its antagonists (Classen et al., 1998, Baraduc
et al., 2004). Accordingly, the effect of TMS on the ballistic task would be very
specific to the generation of a synergic activity burst, and no generic conclusions
about motor learning could be drawn. In contrast to the ballistic task, in the
coordinated arm task a simple modification of CM outflow would not be enough to
improve performance because it does not require a force burst but rather a precisely
coordinated muscle activity (Baraduc et al., 2004). It is possible that the neural bases
of learning coordinated tasks in the motor cortex substantially differs from those
involved in learning simple ballistic tasks.
Second, our single-pulse TMS disruption during blocked practice may not be
as disruptive as the rTMS applied in Muellbacher et al. and Baraduc et al. studies. The
distributed, single-pulse magnetic perturbation applied here may not have accumulated
as much disruptive effect as rTMS to the motor engrams of the trained tasks. Future
investigation of the post-TMS excitability changes induced by single-pulse TMS and
their correlation with task performance indices would help shed light on this issue.
28
The third possibility is that the motor engrams for the coordinated arm task
could be distributed across several cerebral areas. Animal studies of visuomotor
learning have highlighted the role of the supplementary motor area (Wise et al., 1998),
premotor cortex (Mitz et al., 1991, Shen and Alexander, 1997, Wise et al., 1998),
parietal cortex (Eskandar and Assad, 1999), and cerebellum (Ebner and Fu, 1997, Liu
et al., 2003). Alternatively, it has been proposed that learning the dynamics of a
coordinated task leads to the formation of an internal model (Kawato and Wolpert,
1998), which could be stored in the cerebellum (Blakemore et al., 1999, Imamizu et al.,
2000, Liu et al., 2003). Whether these neural substrates directly participate in
encoding or consolidating of motor memory remains to be determined.
Experience-Dependent Neural Plasticity—Support for a General Principle
Our findings support a fundamental principle of the central nervous system
(CNS): plasticity is experience-dependent (Kleim and Jones, 2007a). In particular,
this principle suggests that the nature of the training experience dictates the nature of
the plasticity. Here, practice context characterized the training experience. In the no-
TMS condition, positive plasticity (motor learning) was induced under random
practice conditions. However, when TMS was used to perturb the
encoding/consolidation of that training experience, negative plasticity resulted (poorer
learning). Our experimental design with TMS allowed us to focus on the importance
of the neural implementation of the motor training experience in contrast to the
training experience alone for experience-dependent plasticity. This in turn informs us
about the contribution of the cortical motor system in experience-dependent learning
29
and plasticity. In auditory cortex, the amplitude of the auditory evoked potentials is
modified by musical experience (Kuriki et al., 2006). In the primary somatosensory
cortex, the hand representation is not statically fixed but is dynamically modulated by
practice requirements (Braun et al., 2002). In visual cortex, plasticity is dependent on
the salience of the experience. The pattern of stimulus-induced neuronal activation
depends on the order and interval between successive stimuli (Fu et al., 2002). Our
findings provide a better understanding of how within-session training experience
interacts with the neural implementation of the cortical motor system, and supports the
use of rehabilitative training with variable task practice orders as a tool to improve
brain reorganization and motor learning.
Conclusion
This study provides evidence that the human cortical motor system contributes
to motor skill learning and that the level of its engagement is modulated by within-
session variability of motor practice. With the delayed recall design, we demonstrated
that the cortical motor system is associated with the processing and consolidation of
the trained tasks, especially under a random practice context. This is the first study to
use a specific cortical perturbation to alter the interaction between the practice context
and motor learning. Our results also suggest that the cortical motor system is not a
final storage place for motor memory of coordinated arm movements, but is
functionally relevant for motor memory consolidation.
30
CHAPTER 3
THE ROLE OF HUMAN MOTOR CORTEX IN LEARNING MOVEMENT
KINEMATICS: AN INVESTIGATION USING
TRANSCRANIAL MAGNETIC STIMULATION (TMS) AND
TASK PRACTICE ORDER
Introduction
Neurophysiologic studies in animals and neuroimaging studies in humans have
provided substantial evidence that the cortical motor system (CM) is a putative neural
substrate for motor skill learning (Frost et al., 2000, Plautz et al., 2000, Butefisch et al.,
2004, Pleger et al., 2004, van Mier et al., 2004, Ziemann et al., 2004, Kleim et al.,
2006). There is evidence that the cortical motor system, which includes the premotor
(PMd), primary motor (M1), and somatosensory (S1) cortices, is functionally related
to the planning, encoding, and consolidation of kinematic details of practiced
movements such as amplitude and timing (Classen et al., 1998, Lee et al., 2001,
Butefisch et al., 2004, Crowe et al., 2004, Robertson et al., 2005, Naselaris et al.,
2006b, Townsend et al., 2006). Further, the functional relationship between CM and
motor learning has been shown to depend on the challenge level of the practice
context. For example, greater CM activation is observed when subjects attempt
challenging motor tasks or are trained in a variable practice condition (Demiralp et al.,
1990, Mitz et al., 1991, Thach, 1997, Zhang et al., 1997, Sergio and Kalaska, 1998,
Chung et al., 2000, Schubotz and von Cramon, 2002, Ben-Shaul et al., 2004, Gorbet et
31
al., 2004, Hatfield et al., 2004, Lewis et al., 2004, Seidler et al., 2004, Aoki et al., 2005,
Meister et al., 2005). In a previous study, we demonstrated a correlate between CM
activity and practice context by manipulating task practice order. A magnetic
perturbation centered over the motor cortex during practice attenuated motor learning
under random-order practice, but not blocked-order practice (Lin et al., 2007b). These
findings suggest that CM activity is evoked particularly during random-order practice.
Given that random-order practice is a valid motor learning paradigm to evoke CM
activity, it was chosen for this study.
The primary aim of this study was to investigate the specific role of the cortical
motor system (CM) in motor learning of an arm task that has specific amplitude and
timing requirements. Given that CM is more engaged during random-order practice,
we reasoned that if CM has a specific role in processing kinematic requirements of the
trained tasks, this role should be evident under a random practice condition. To test
this hypothesis, transcranial magnetic stimulation (TMS) was employed as
perturbation to interfere with CM activity during practice of three goal-directed arm
movements (Flitman et al., 1998a, Hallett, 2000, Boroojerdi et al., 2001, Ziemann,
2004). Single TMS pulse was synchronized to each inter-trial interval, a time in
which kinematics planning and encoding are thought to occur (Lee and Magill, 1983,
Shea and Zimny, 1983, Butefisch et al., 2004). If CM is a functional substrate for
processing the task-relevant timing and/or amplitude, perturbation over CM would
interfere with learning of these kinematic details under the random practice condition
in which CM is particularly engaged.
32
Materials and Methods
Subjects
Thirty one healthy, right-handed volunteers (18 - 38 years) who were naive to
the task were recruited (Table 2). There were no significant differences across
experimental groups for age and gender. All participants gave written informed
consent and participated in the study under a protocol approved by the Institutional
Review Board of the University of Southern California Health Sciences Campus.
Exclusion criteria were all contraindications to receiving transcranial magnetic
stimulation (Wassermann, 1998) including: the presence of a pacemaker, metal in head,
pregnancy, any neurological disorder, current use of stimulants or medications known
to lower seizure threshold, and personal or family history of a seizure disorder.
Instrumentation and Task
A lightweight lever affixed to a frictionless vertical axle was attached to a table
and positioned parallel to the floor. A handle at the end of the lever was adjusted to
accommodate the participant’s forearm. A linear potentiometer was attached to the
transducer at the base of the vertical axle. Signals from the transducer were converted
to a digital signal and sampled at 1000 Hz. A LabView-based (Labview, National
Instruments) custom-made software program (Weekley, 2004) was applied to
manipulate the movement trajectory, the timeline of each event, and data storage for
each trial. The motor task was to move the lever in order to replicate a template
movement trajectory, which was displayed on the computer monitor before each trial
(Fig. 1). The goal was to learn three discrete arm movement trajectories composed of
33
two extension-flexion reversal actions (Fig. 1). The three trajectories were
differentiated by the timing and amplitude of the reversal actions. The three
movement trajectories were presented with a quasi-random order (i.e., C-A-B-, A-B-
C-, B-C-A-, if each letter represents a template trajectory). During the practice phase,
the participants were presented with feedback after each movement trial. The
feedback included 1) an overall error score, representing the difference between the
template movement trajectory and the participant’s response and 2) a graphic
representation of the participant’s response superimposed on the template movement
trajectory. All participants received written instructions and additional verbal
information about the experiment. The timeline of each event during a single trial is
illustrated in Figure 2. Participants were allowed to perform each movement up to
five seconds. The time to complete the movement was variable and depended on how
accurately the participant performed, whereas the other interval times (e.g., feedback
delay) were fixed.
Experimental Design:
Participants were randomized to one of the three practice conditions, derived
from each of three Stimulation Conditions: no-TMS Control, TMS, and Sham-TMS.
TMS pulses were delivered to the primary motor cortex contralateral to the moving
arm. The sham TMS condition (i.e., coil applied to same location, normal clicking
noise apparent, but no magnetic pulse applied) was implemented to control for the
non-specific effects of TMS including coil noise and scalp sensation.
34
Experimental Procedure
Since the learning performance distinction is a well-recognized phenomenon
in learning and memory research, we employed a recall-test paradigm in order to
measure the relatively permanent effects on motor performance (Shea and Morgan,
1979, Hall and Magill, 1995, Goodwin and Meeuwsen, 1996, Lee et al., 1997, Cahill
et al., 2001, Meira and Tani, 2001, Giuffrida et al., 2002, Perez et al., 2005, Schmidt
and Lee, 2005, Wright et al., 2005). Participants came for two consecutive days with
practice phases on day 1 and retention phase on day 2. Participants practiced three
sets of 48 trials during the practice phase, resulting in a total of 144 practice trials.
Participants performed 24 recall accuracy trials in retention phase. Trials in the
retention phase were administered without feedback, TMS or Sham-TMS.
Magnetic Stimulation
Magnetic pulses were delivered by a magnetoelectric stimulator (MES-10,
Cadwell Laboratories, Kennewick, WK) through a circular magnetic coil (9.0 cm
diameter). In considering the duration of the practice period (144 trials which lasted
for around 90 minutes), a circular coil was chosen rather than a figure-8 as there are
less overheating problems with this coil. The coil was placed tangentially to the scalp
in a direction such that monophasic pulses induced a current flowing antero-
posteriorly perpendicular to the central sulcus (Werhahn et al., 1994, Kaneko et al.,
1996). To center our stimulation over the motor cortex, we placed one edge of the coil
on the subject’s scalp and identified an optimal location where stimuli consistently
elicited a maximal motor evoked potential (MEP) in the contralateral biceps brachii
35
muscle (Kamen, 2004). Motor threshold was defined as the minimum TMS intensity
to the nearest 1% of maximum stimulator output that evoked a MEP ≥ 50 μV in at
least 5 out of 10 trials in the target muscle (Rossini et al., 1994). TMS pulses were
only applied during the practice phase and synchronized to the onset of each inter-trial
interval (Fig. 2). This interval was chosen as a critical period for cognitive processes
during practice when information from the previous trial might be used in the
preparation for the next trial (Pascual-Leone et al., 1992b, Berardelli et al., 1994,
Taylor et al., 1995, Ziemann et al., 2004). A suprathreshold TMS intensity (120% of
resting motor threshold) was chosen in order to produce a disruptive influence on
cortical activity centered on and around the motor cortex.
Dependent Measures
Accuracy in movement timing and movement amplitude was quantified across
acquisition and retention phases. Accuracy in movement timing was quantified using
absolute error in movement time (AE_MT, Fig. 5, next page). Accuracy in movement
amplitude was quantified using the sum of absolute error in movement amplitude
derived from the three reversal points within each movement pattern (AE_SumAmp,
Fig. 5, next page). Absolute amplitude error at each reversal point was calculated by
placing an event-marker on each reversal point using a customized MatLab based
program (Wu, 2004). Twelve trial blocks were combined for calculation of average
AE_MT and AE_SumAmp. Data of the retention phases were calculated after
collapsing across two testing conditions.
36
Statistical Analysis
For the acquisition phase, average performance errors (timing error and
amplitude error) were submitted to a 3 Stimulation Condition (no TMS Control, TMS,
and Sham TMS) × 2 Practice Order (Blocked and Random) × 12 Block (block 1 to 12)
analysis of variance with repeated measures on the last factor. A 3 (Stimulation
Condition) × 2 (Practice Order) ANOVA was calculated for the retention phase.
Effect size (ES) was used to estimate the magnitude of between group differences and
Figure 5. Accuracy in movement timing was quantified using absolute error in
movement time (AE_MT). Accuracy in movement amplitude was
quantified using the sum of absolute error in movement amplitude
derived from the three reversal points within each movement trajectory
(AE_SumAmp). AE-ampPeak1: absolute value of amplitude error at the
peak1; AE-ampTrough: absolute value of amplitude error at the trough;
AE-ampPeak2: absolute value of amplitude error at the peak2.
37
can be meaningful for studies where sample sizes are small. ES is reported according
to established criteria as small (< .41), medium (.41-.70), or large (> .70) (Thomas et
al., 1991). When appropriate, the ANOVAs were followed by a simple main effect
test and Post hoc Tukey test to control for inflated Type 1 error. For all statistical tests,
significance level was set at p < 0.05. SPSS 13.0 (SPSS Inc., Chicago, IL) statistical
software was used for all statistical analysis.
Results
Group mean comparisons for age, gender, and resting motor threshold are
summarized in Table 2. There were no significant differences across experimental
groups for age and gender. The resting motor threshold prior to motor practice did not
differ significantly between the Blocked-TMS and Random-TMS groups.
Table 2. Baseline demographic information
Stimulation
Condition
No TMS (Control) TMS Sham TMS
Age 28.4±1.4 28.4±2.1 24.6±1.1
Gender 6M, 4F 3M, 7F 5M, 6F
M: male, F: female
Acquisition phase
Timing Error (AE_MT)
The ability to control movement timing was not different among groups in
baseline (block 1) (p = 0.539, Fig. 6, block 1). Each group showed reduction of
timing error over the practice phase that resulted in a significant main effect for
38
Practice Block (p < 0.001, Fig. 6). The improvement of timing error was no
influenced by Stimulation Condition (p= 0.228).
Amplitude Error (AE_SumAmp)
The ability to control movement amplitude was not different among groups in
baseline (p = 0.766, Fig. 7, block 1). Each group showed a reduction of amplitude
error over the practice phase that resulted in a significant main effect for Practice
Block (p < 0.001, Fig. 7). The improvement of amplitude error was no influenced by
Stimulation Condition (p= 0.104).
Retention Phases
Timing Error (AE_MT)
Stimulation Condition influenced the recall accuracy of the movement timing
(p= .024, Fig. 6, block 13). Magnetic stimulation during practice phase deteriorated
the recall accuracy of movement timing for the TMS group (higher timing error, Fig.
6). Sham-TMS did not affect the recall accuracy since the Sham-TMS group
performed similarly to the no-TMS Control group. The timing error of the TMS group
was higher than that of the Control group (p = 0.032) and the Sham-TMS group (p =
0.067, effect size= 0.9), suggesting that the detrimental effect of magnetic stimulation
on the recall accuracy of movement timing was mediated through direct cortical
perturbation.
Amplitude Error (AE_SumAmp)
Stimulation Condition did not significantly influence the recall accuracy of the
movement amplitude (p= .105, Fig. 7, block 13) though visual inspection suggests a
39
trend that motor practice with TMS and Sham TMS deteriorated this recall accuracy.
The effect size of group differences between Random-Control and Random-TMS
(effect size= 0.94, Fig. 7), and between Random-Control and Random-Sham TMS
(effect size= 0.89) were high, suggesting that TMS and Sham TMS during practice
phase were detrimental to later recall of movement amplitude.
Discussion
The cortical motor system (CM) has been shown to encode kinematic details of
the practiced movements as part of its role in motor skill learning (Classen et al., 1998,
Butefisch et al., 2000, Butefisch et al., 2004, Townsend et al., 2006). The present
Figure 6. Timing errors for each experimental group across acquisition and
retention phases. The acquisition phase includes Block 1 to 12;
Block 13 is the second-day retention phase. The timing error of the
retention phase was the average error of two testing conditions (i.e.
blocked and random). * indicates the group differences are
statistically significant. Error bars are standard errors of mean.
40
study applied transcranial magnetic stimulation (TMS) to perturb corticomotor activity
while participants practiced three arm tasks each with specific spatial and temporal
requirements.
To ensure the practice condition evokes CM activity, participants practiced
motor tasks under a random-order condition, a condition in which CM is particularly
engaged (Lin et al., 2007b). We hypothesized that if CM is a functional substrate for
processing task-relevant timing and/or amplitude, perturbation over CM would
Figure 7. Amplitude error in degree of the six experimental groups across
acquisition and two retention phases. The acquisition phase includes
Block 1 to 12; Blocks 13 is the immediate retention phase; Blocks 14
is the delayed retention phase. The RMSE of retention tests are the
average after clasping two testing conditions (i.e. blocked and
random). Error bars are standard errors of mean.
41
interfere with learning of these kinematic details. The main finding was that TMS
perturbation interfered with learning of movement timing, but not amplitude.
However, this disrupting effect was not observed in the sham TMS condition. The
results will be discussed within the context of CM’s role in processing movement
timing and amplitude, followed by some speculation on how the results may provide
insight into a well-known attribute of the cortical motor system in motor learning.
Cortical Motor System and Learning of Movement Timing
We demonstrated that during motor practice, TMS perturbation over the
cortical motor system (CM) deteriorated the learning of movement timing (Fig. 6),
suggesting that CM is a neural substrate for processing temporal information during
motor skill learning. Interestingly, this disruptive effect of TMS to the learning of
movement timing was not robust until the retention phase (block 13, Fig. 6),
suggesting that not all changes during the practice phase represent permanent changes
in motor behavior (the definition of learning) (Cahill et al., 2001, Schmidt and Lee,
2005), and provides support for the known Learning Performance Distinction in the
skill learning literature. It also informs us about the importance of evaluating
“learning” with a retention test design. One might not identify CM’s role in learning
the movement timing if motor learning was evaluated merely by performance during
the practice phase.
We presume that the greater CM activity for processing movement timing
during motor practice is the target of the TMS disruptive effect. CM is inhibited by
the TMS pulses from processing temporal information for subsequent motor memory
42
formation. Previous studies suggest that motor memory formation is susceptible to
interference by a focal lesion or competing tasks until the memory traces have been
consolidated (Muellbacher et al., 2002, Krakauer and Shadmehr, 2006). In the present
experiment, TMS during the inter-trial interval created a virtual “lesion” that
ultimately interfered with the consolidation of the processed task-timing information
resulting in poor retention performance.
There are at least two possible ways TMS could modulate learning of
movement timing: (i) cortical magnetic induction or (ii) non-specific elements during
the stimulation procedure, such as the clicking noise when coil discharged. The fact
that the retention performance were similar between the Sham-TMS and the no-TMS
Control groups, and both were better than the TMS group suggests that, TMS
interfered with learning of movement timing occurs through direct perturbation of
cortical motor system.
Alternatively, the function of CM in processing movement timing could be
disrupted perhaps in ways other than cortical stimulation. For example, the peripheral
muscle twitches induced by supra-threshold TMS could potentially "perturb" motor
learning through somatosensory input. It is also possible that the learning of
movement timing could be disrupted in ways other than direct cortical perturbation to
CM. In fact, there is little evidence that the cortical motor system has a direct
contribution in processing movement timing. We speculate that TMS perturbation
centered on and around motor cortex could have indirectly caused the learning deficit
in timing by disrupting the connections/synchronization between the motor cortex and
43
other neural substrates known for processing movement timing, such as the
supplementary motor area, posterior parietal area, and cerebellum (Thickbroom et al.,
2000, Macar et al., 2006, Martin et al., 2006). Alternatively, the TMS pulses could
have influence the dorsal premotor area and the supplementary motor area. Those
alternative explanations suggest that the cortical motor system might not directly
process timing information, but is either i) functionally connected to the neural
substrates known for processing timing information, or ii) functionally relevant to the
consolidation of the processed timing information.
Cortical Motor System and Learning of Movement Amplitude
In this study, cortical perturbation did not significantly interfere with the
learning of movement amplitude (Fig. 7, block 13). One interpretation is that CM
does not process movement amplitude. Consequently TMS-induced CM
perturbation would not cause deterioration in amplitude learning. This interpretation
is however unlikely given that CM has been associated with planning, adjusting, and
encoding task-relevant spatial information (Crowe et al., 2004, Diedrichsen et al.,
2005, Liu et al., 2005, Monfils et al., 2005, Chouinard and Paus, 2006, Coxon et al.,
2006, Lee and van Donkelaar, 2006, Naselaris et al., 2006a). Previous studies which
applied neurophysiology and neuroimaging methods have typically identified a role
for CM in processing movement amplitude. Our results suggest that this processing
may be modulated by the strategy participants adopt to improve task performance.
For example, in this and previous studies conducted in our laboratory, the best
strategy for participants to reduce error in the lever arm task is to improve movement
44
timing (Lin et al., 2007a). Therefore, it is conceivable that CM is less tuned to
processing movement amplitude since the behavioral preference for this task is to
optimize movement timing. Further work requiring the learning of a spatially-biased
task could be used to address this hypothesis.
Experience-specific Neural Plasticity underlies Motor Learning
Our findings may be explained by experience-specific neuroplasticity of the
motor cortex. (Kleim and Jones, 2007b). In particular, experience-specific
neuroplasticity suggests that the nature of the training experience dictates the nature
of the plasticity. In other words, the relevant information to fulfill the requirements
of the motor task (e.g., timing information for the lever arm task) may be what is
prioritized by the cortical motor system. Here, in the no-TMS condition, participants
showed dramatic improvement in movement timing but less improvement in
amplitude (Figs, 6 and 7), suggesting that participants optimized performance of the
practiced tasks mainly by improving movement timing. Therefore, if the “behavioral
requirement” for processing movement timing is greater than that of the movement
amplitude, it is not surprising that CM would tune to the timing feature in order to
fulfill this behavioral constraint. Our experimental design with TMS provides a
better understanding of how behavior requirement interacts with the neural
implementation of the cortical motor system.
One potential confound in the current study is that the three template
trajectories differed in amplitude but not in movement time. One could argue that
participants primarily learned movement timing simply because the repetition of the
45
movement time was more than that of the movement amplitude. In order to rule out
this potential confound, we grouped the data by each template trajectory and
demonstrated that the participants showed no apparent transfer of movement time
from one template to the next. Despite having reduced movement time errors from
approximately 300 ms to less than 100 ms by the end of practice with one template,
movement time errors for the next template jumped right back up to the 300 ms
range, suggesting that the participants had no awareness that the new template had
the same duration as the previous one. This suggests that the findings of our study
are not confounded by the more frequent repetition of movement time compared
with amplitude across three template trajectories.
In summary, we demonstrated that the cortical motor system is a functional
substrate for processing and consolidating task-relevant timing information of the
trained tasks. Our results also suggest that the adult motor cortex may prioritize the
processing of selected kinematic details to fulfill specific task constraints. Future
studies should also focus on determining whether the cortical motor system directly
processes movement kinematics as its role in motor learning or it simply serves as a
“proxy” in the neural circuitry for motor learning. From the clinical point of view,
this study provides a mechanistic foundation for developing cortical-location
specific therapies where brain is directly influenced, such as repetitive TMS (rTMS),
transcranial direct current stimulation (tDCS), or cortical electrical stimulation by a
stimulator directly implanted in the motor cortex.
46
CHAPTER 4
THE CONTEXTUAL INTERFERENCE EFFECT: ELABORATION-
DISTINCTIVENESS OR FORGETTING-RECONSTRUCTION? A POST-HOC
ANALYSIS OF TMS-INDUCED EFFECTS ON MOTOR LEARNING
Introduction
It is well established that, when learning multiple tasks or multiple variants of
the same task, random practice leads to better retention than blocked practice, known
as the Contextual Interference (CI) Effect (Shea and Morgan, 1979, Lee and Magill,
1983). It has long been of interest to motor learning theorists to understand the
information processing mechanisms underlying this phenomenon. Two principal
explanations for the CI effect are the (I) elaborative processing and (2) forgetting-
reconstruction hypotheses. The elaborative processing hypothesis holds that random
practice forces the learner into more elaborate processing such as inter-task
comparisons and embellishment of task-relevant information (Shea and Morgan, 1979,
Shea and Zimny, 1983, Shea and Zimny, 1988, Shea and Titzer, 1993). More
elaborate information processes are thought to result in more comprehensive and
readily retrievable memory traces. According to the elaborative processing hypothesis,
the interval prior to the next movement trial (i.e., inter-trial interval) is a likely time
during which inter-task comparisons are made and the elaboration occurs.
The forgetting reconstruction hypothesis suggests that a previously constructed
action plan is more likely to be available in working memory during blocked practice,
47
since the same task is practiced repeatedly (Lee and Magill, 1983, Lee and Magill,
1985). Random practice however, forces the learner to abandon the previously
constructed action plan because he or she has to perform a different task on the next
trial. The forgetting that ensues requires the learner to actively reconstruct the action
plan the next time the initial task is practiced, resulting in a stronger memory
representation of the practiced tasks (Lee and Magill, 1983, Lee and Magill, 1985).
As with the elaborative processing hypothesis, forgetting and reconstruction of the
action plan are likely to occur in the interval prior to the next movement trial.
The present experiment attempted to distinguish which of the two hypotheses
better account for the CI effect by applying a transcranial magnetic stimulation (TMS)
pulse delivered to the motor representation of the primary muscles of the trained tasks.
TMS is a technique that can enhance or inhibit central reorganization (Ziemann et al.,
1998) and information processing (Flitman et al., 1998b, Boroojerdi et al., 2001) by
means of noninvasive focal stimulation of the human brain (Hallett, 2000). Unique to
the TMS technique is its temporal resolution (in millisecond scale), which in turn
allows researchers to directly perturb hypothetical information processes evoked
during motor control and learning. This technique has been applied to enhance or
inhibit information processes during skill acquisition (Pascual-Leone et al., 1992a,
Berardelli et al., 1994, Taylor et al., 1995, Ziemann et al., 2004).
Single disruptive TMS pulses were synchronized to each inter-trial interval to
modulate information processes of a blocked and random practice condition.
Performance during retention was compared to the control conditions in which no
48
TMS pulses were applied. This experimental manipulation should lead to one of two
most-likely results, depending on whether elaborative processing or forgetting-
reconstruction provides a better explanation for the contextual interference effect
(Figure 1). If elaborative processing accounts for the CI effect, then the advantage of
random practice should be diminished by the TMS, since TMS perturbs elaborative
processing during the inter-trial interval (Figure 1A, left). Motor learning under
blocked practice condition will not be affected since little or no elaborative processes
are evoked during blocked practice (Fig. 1A, right). On the other hand, if forgetting
and reconstruction accounts for the CI effect, then blocked practice should become
more like random practice, since TMS applied during the inter-trial interval ensues
reconstruction of the action plan even during blocked practice in which the same task
is practiced repeatedly (Figure 1B, right). Motor learning under random practice is not
expected to be significantly enhanced by additional TMS perturbation, since forgetting
and reconstruction of action plans have been evoked by random practice (Fig. 1B, left).
Materials and Methods
Subjects
Forty healthy volunteers (18 - 38 years) who were naive to the task and the
purpose of the experiment were recruited. All participants gave written informed
consent and participated in the study under a protocol approved by the Institutional
Review Board of the University of Southern California Health Sciences Campus.
Participants were excluded from the study if there were contraindications for
transcranial magnetic stimulation (Wassermann, 1998) including: the presence of a
49
pacemaker, metal in the head, pregnancy, any neurological disorder, current use of
stimulants or medications known to lower seizure threshold, and personal or family
history of a seizure disorder.
Instrumentation and Task
A lightweight lever affixed to a frictionless vertical axle was attached to a table
and positioned parallel to the floor. A handle at the end of the lever was adjusted to
accommodate the length of participant’s forearm. A linear potentiometer was attached
to a transducer at the base of the vertical axle. Signals from the transducer were
converted to a digital signal and sampled at 1000 Hz. A LabView-based (Labview,
Figure 8. Predicted results of the experimental manipulation based on Elaborative
processing hypothesis (A) and Forgetting reconstruction hypothesis (B).
TMS: disruptive transcranial magnetic stimulation pulses
50
National Instruments) custom-made software program (Weekley, 2004) was applied to
manipulate the movement pattern, the timeline of each event, and data storage for each
trial. The participant’s task was to move the lever with their dominant arm at the
correct speed and distance to replicate a goal movement pattern that was displayed on
the computer monitor before each trial (Figure 1). The participant was required to
learn three arm patterns each composed of two elbow extension-flexion reversal
movements (Figure 1).
During the acquisition phase of the experiment, the participants were presented
with feedback after each movement trial. The feedback included 1) an overall error
score, root mean square error (RMSE), representing the difference between the goal
movement pattern and the participants response and 2) a graphic representation of the
participant’s response superimposed on the goal movement pattern. All participants
received written instructions and additional verbal information about the experiment.
The timeline of each event of a single trial is illustrated in Figure 2.
Experimental Design and Procedure
Participants were randomized to one of four practice conditions: no TMS
control (Blocked, Random), and TMS (Blocked-TMS, Random-TMS). There were no
significant group differences for age, gender, or baseline motor performance (block 1,
Figure 9). The resting motor threshold prior to motor practice did not differ
significantly between the Blocked-TMS and Random-TMS groups.
Testing took place over two consecutive days with acquisition and immediate
retention phases on day 1 and a delayed retention phase on day 2. Participants
51
practiced three sets of 48 trials (48 trials per goal movement pattern) during the
acquisition phase, resulting in a total of 144 practice trials. The retention tests
consisted of no-feedback and no-TMS trials. To ensure that the condition of retention
test did not favor either practice order, participants were tested by both blocked and
random order retention tests with 12 trials in each test. Retention test order was
counterbalanced across participants.
Magnetic Stimulation
TMS pulses were applied to the motor cortex contralateral to the moving arm
by a magnetoelectric stimulator (Cadwell Laboratories, Kennewick, WK) through a
focal circular magnetic coil (diameter: 9.0 cm). The coil was placed tangentially to
the scalp in a direction such that monophasic pulses induced a current perpendicular to
the central sulcus in an optimal position for activating the corticospinal tract trans-
synaptically (Werhahn et al., 1994, Kaneko et al., 1996). The coil was positioned so
that stimuli consistently elicited isolated elbow-flexion movements and produced a
motor evoked potential (MEP) in the contralateral biceps brachii muscle (Kamen,
2004). Motor threshold was defined as the minimum TMS intensity that evoked a
MEP in the target muscle with an amplitude greater than or equal to 50 μV in at least 5
out of 10 trials (Rossini et al., 1994). Motor threshold was determined to the nearest
1% of maximum stimulator output. We chose to apply TMS with an intensity that was
120% of the resting motor threshold to induce an inhibitory effect on the motor cortex
(Pascual-Leone et al., 1992b, Berardelli et al., 1994, Taylor et al., 1995, Ziemann et al.,
52
2004). A single TMS pulse was synchronized to the onset of each inter-trial interval
during the acquisition phase (Figure 2).
Dependent Measures
Overall performance accuracy was quantified separately for acquisition and
retention phases using root mean square error (RMSE) (Wulf et al., 1993). RMSE is
the mean difference between the goal movement pattern and the participant’s response
calculated over the participant’s total movement time. RMSE of each trial was
calculated after synchronizing the onset of the goal pattern with the onset of the
participant’s movement. For display of the mean RMSE across acquisition phase,
RMSE was averaged across consecutive 12 trials. For display of the mean RMSE
during the immediate and delayed retention phases, RMSE was averaged across
blocked and random retention tests with 12 trials in each test. Another way to asses
the memory strength of the practiced tasks is to compare the performance after 24
hours (delayed retention) to the performance immediately after practice (immediate
retention). A group analysis was performed for this day2-day1 decrement, defined as
the RMSE changed score between immediate and delayed retention phases (i.e., mean
RMSE of the delayed retention phase – mean RMSE of the immediate retention phase).
Statistical Analysis
For the acquisition data, mean RMSE across the 144 acquisition trials was
submitted to a 2 Stimulation Condition (noTMS, TMS) × 2 Practice Order (Blocked,
Random) × 12 Acquisition Block analysis of variance (ANOVA) with repeated
measures for acquisition block. A 2 (Stimulation Condition) × 2 (Practice Order)
53
ANOVA was conducted separately for 1) the mean RMSE of immediate and delayed
retention phases; and 2) the decrement from day 1 to day 2. When appropriate, and
for determining the locus of any significant interaction, the ANOVAs were followed
by a simple main effect test and Post hoc Tukey test to control for inflated Type 1
error. Effect size (ES) was used to estimate the magnitude of between group
differences. ES is reported according to established criteria as small (< 0.41), medium
(0.41 - 0.70), or large (> 0.70) (Thomas et al., 1991). For all statistical tests,
significance level was set at p < 0.05. SPSS 13.0 (SPSS Inc., Chicago, IL) statistical
software was used for all statistical analysis.
Results
Acquisition phase
TMS perturbation influenced blocked and random practice differentially
during the acquisition phase (Figure 9, blocks 1 to 12). For participants who practiced
the three tasks in blocked order condition, the presence of a TMS perturbation during
the inter-trial interval actually benefited performance (average across 144 acquisition
trials) compared to participants who practiced without TMS (Figure 9, left). In
contrast, for participants who practiced the tasks in random order condition, TMS
worsened acquisition performance compared to the no-TMS condition (Figure 9, right).
This differential effect of Stimulation Condition on Practice Order is associated with a
significant Stimulation Condition by Practice Order interaction (p = .012, Figure 9).
Post hoc tests showed no difference between the two control groups (Blocked-Control
and Random-Control) but a significant difference between the two TMS groups
54
(Blocked-TMS and Random-TMS), p = .005, Figure 10A. In general, there was a
trend in which TMS perturbation enhanced acquisition performance of the Blocked-
TMS group compared to that of the Blocked-Control group (p = .08, Figure 10A). But
TMS deteriorated performance of the Random-TMS group compare to that of the
Random-Control group (p = .07, Figure 10A).
Immediate Retention
For participants who practiced the tasks in both blocked and random conditions,
motor performance during immediate retention was not affected by TMS perturbation
Figure 9. Mean root mean square error (RMSE) of the four experimental
groups during acquisition (block 1 to 12), immediate (block 13), and
delayed retention (block 14) phases. Blocked: blocked practice,
Random: random practice. * indicates the group differences are
statistically significant.
55
(Figure 9, block 13). Motor practice with TMS perturbation did not differentially
affect the immediate retention performance of blocked and random practice, resulting
in no significant Stimulation Condition by Practice Order interaction.
Figure 10. (A) Mean root mean square error (RMSE) of the acquisition phase. The
X axis is the Practice order. (B) RMSE of the delayed retention phase. *
indicates that the group differences are statistically significant.
(A)
(B)
56
Delayed Retention
When subjects were re-tested a day after acquisition without TMS, the
contextual interference effect in motor skill learning was demonstrated. In the control
no-TMS groups that had not been exposed to TMS during acquisition, the subjects
who trained with a random order performed better than the subjects who trained with a
blocked order (p = .01, Figures 9 and 10B). The effect of TMS perturbation during
acquisition to the retention performance depended on whether the subjects practiced
with the blocked or random practice orders (Figures 9 and 10B). TMS perturbation
during the acquisition phase did not influence retention performance of the Blocked-
TMS group (p= .40, Figure 10B) but led to increased errors for the Random-TMS
group (p = .01, Figure 10B). This differential effect of TMS perturbation on blocked
and random practice was associated with a significant Stimulation Condition by
Practice Order interaction (p= .014).
Decrement from Day1 to Day2
Another way to assess the memory strength of the practiced tasks is to measure
the decrement from immediately after practice (immediate retention) to 24 hours later
(delayed retention). Consistent with the average RMSE during the delayed retention
phase, for subjects who practiced with a random order, a TMS perturbation led to
significant decrement in performance from day1 to day2 compared to the subjects who
had no TMS perturbation (p = .01, Figure 11). TMS perturbation did not influence
this decrement for the blocked practice group (p= .66). This differential effect of TMS
57
perturbation on “decrement” for the blocked and random practice is associated with a
significant Stimulation Condition by Practice Order interaction (p= .03).
Discussion
The purpose of the present study was to explore the viability of the two
principal hypotheses of the contextual interference effect in motor skill learning.
Transcranial magnetic stimulation (TMS) was applied to perturb information
processes of each inter-trial interval, in which task elaboration and action plan
reconstruction are more likely operational. If the extensive task elaboration is
essential to enhance motor learning, as predicted by the elaborative processing
Figure 11. Mean decrement in performance from day1 to day2, defined as the mean
RMSE changes between the immediate and delayed retention phases.
The X axis is the Practice order. * indicates that the group differences are
statistically significant. Error bars are standard errors of mean.
58
hypothesis, perturbing the task elaboration would interfere with motor learning under
random practice but not blocked practice. On the other hand, if reconstruction of
action plan benefits learning, as suggested by the forgetting reconstruction hypothesis,
additional perturbation during motor practice should encourage action plan
reconstruction therefore promote motor learning under blocked practice. Our results
support the elaborative processing hypothesis by showing that disruption of inter-trial
processes deteriorated the learning benefit of random practice. The forgetting
reconstruction hypothesis was not supported as the additional perturbation during
acquisition phase did not benefit motor learning for the blocked practice (trend
without significance).
The contextual interference effect was replicated in the learning of fast discrete
arm tasks (Shea and Morgan, 1979, Hall and Magill, 1995, Goodwin and Meeuwsen,
1996, Lee et al., 1997, Meira and Tani, 2001, Giuffrida et al., 2002, Perez et al., 2005,
Wright et al., 2005). Without TMS perturbation, participants who practiced in the
random order condition demonstrated better retention performance than did those who
practiced in the blocked order condition (Figure 10). However, when information
processes during each inter-trial interval were perturbed by a TMS pulse, this learning
benefit was abolished: the Random-TMS group performed with greater errors
compared with the Random-Control group during the acquisition phase, and
demonstrated even greater deficits in retention performance (Figures 9-10). In
contrast, TMS perturbation during blocked practice did not deteriorate motor
performance and learning. This pattern of asymmetric double dissociation during
59
delayed retention (i.e., TMS perturbation interfered with retention performance of
random practice but not blocked practice) supports the elaborative processing
hypothesis. Deeper information processes evoked by random practice during
acquisition phase, such as task elaboration, provides long term benefit for motor
learning. Perturbation of these processes abolishes the contextual interference effect.
Our results did not follow the prediction of the forgetting reconstruction
hypothesis, which predicts greater perturbation during acquisition phase would
promote learning. Motor learning (i.e., retention) under blocked practice was not
significantly enhanced by the additional disruption during acquisition (trend without
significance, Figure 9), suggesting that reconstruction of the action plan might be
beneficial, but not sufficient to explain the benefit of random practice for motor
learning. Another line of evidence in opposition to the forgetting reconstruction
hypothesis is the data of the acquisition phase. The forgetting reconstruction
hypothesis claims that random practice causes forgetting, which is evident by the
detrimental acquisition performance (Lee and Magill, 1983). However, this is not the
case in our Blocked practice groups when TMS perturbation was introduced (Figure 9,
blocks 1-12). Instead, the acquisition performance of the blocked practice was
actually enhanced by the TMS disruption (Figure 10A), suggesting that TMS
perturbation influenced performance of blocked practice through mechanisms other
than forgetting-induced reconstruction. Post-experiment interviews by Shea and
Zimny (1983), indicated that participants in the blocked condition tended to run the
movements off without much thought, more or less “automatically” (Shea and Zimny,
60
1983), we therefore speculate that the additional TMS perturbation during blocked
practice may necessitate a reallocation of attentional resources to the motor tasks,
resulting in better performance during acquisition and delayed retention. This
interpretation is supported by the more apparent performance improvement (Figure 9)
during early acquisition phase and the more accurate performance during the delayed
retention phase of the Blocked-TMS group compared to its no TMS control
counterpart (Figure 9).
Given the previous argument, the increased attentional demand of random
practice plus TMS perturbation should enhance performance even more compared
with random practice alone. However, TMS disruption during random order practice
was not beneficial but actually detrimental for motor performance and learning (Figure
9, right). If greater perturbation during acquisition necessitates a reallocation of
attentional resources, our data suggest that there is a desirable level of attention which
leads to the optimal learning outcome (Kahneman, 1973). Though additional attention
demand as a result of TMS perturbation enhanced performance of the blocked practice,
this increasing demand worsened motor learning for an initially demanding condition
such as random order practice (Li and Wright, 2000).
To confirm the interpretation that TMS perturbation during blocked practice
necessitates a reallocation of attentional resources to motor practice, we added a
Blocked-Interpolation group as a comparison. This Interpolation condition was
designed to induce distraction during blocked practice similar to what we believed was
occurring with TMS. We found that the performance of this Blocked-Interpolation
61
group was very similar to the Blocked-TMS group throughout the acquisition and
retention phases, suggesting that “attention” can be a complementary explanation for
the contextual interference effect. In summary, the forgetting reconstruction
hypothesis does not adequately explain the results in our study. Other information
processes, such as attention and task elaboration should be considered to fully explain
the benefit of random practice.
In addition to providing support for the elaboration processing hypothesis for
the contextual interference effect, we have introduced an experimental approach
(transcranial magnetic stimulation technique) to probe information processes during
motor learning. Unique to the transcranial magnetic stimulation paradigm is its high
temporal resolution (in millisecond scale), which makes it a valuable tool for
disrupting key processes that occur at specific points in time during the learning of a
motor skill.
62
CHAPTER 5
SUMMARY AND GENERAL DISCUSSION
Introduction
The main purpose of this dissertation was to examine the neuromotor
mechanism which implements the contextual interference (CI) effect in motor skill
learning. An important operational assumption is that random practice of motor skills
is more challenging than blocked practice (Li and Wright, 2000). My goal was to
determine whether the CI effect can be explained by different brain activity when
learners practice motor skills under a blocked compared to a quasi-random practice
condition. Specifically, we delivered transcranial magnetic stimulation (TMS) pulses
centered and around the motor cortex to create a perturbation of the cortical motor
system (CM) processing during the practice phase. The effect of perturbation on
motor learning was determined by examining the recall performance accuracy during a
delayed retention phase. CM was chosen as the target for TMS for two reasons. First,
there is substantial evidence from neurophysiologic and neuroimaging studies showing
that CM exhibits significant engagement during motor skill learning particularly when
task demands are challenging or novel (Mitz et al., 1991, Grafton et al., 1992,
Kawashima et al., 1994, Elbert et al., 1995, Nudo et al., 1996, Thach, 1997, Classen et
al., 1998, Cohen et al., 1998, Karni et al., 1998, Kleim et al., 1998, van Mier et al.,
1998, Classen et al., 1999, Hess et al., 1999, Hund-Georgiadis and von Cramon, 1999,
Liepert et al., 1999, Frost et al., 2000, Plautz et al., 2000, Ben-Shaul et al., 2004,
63
Butefisch et al., 2004, Hatfield et al., 2004, Pleger et al., 2004, van Mier et al., 2004,
Ziemann et al., 2004, Kleim et al., 2006). Second, when the motor representations of
the motor cortex are stimulated by supra-threshold magnetic pulses, an evoked muscle
response is observed in the projecting muscles. Therefore, reliably evoked muscle
responses were used to verify the stimulating location of the magnetic pulses. This
dissertation included two between-subject factors in the experimental design. The two
factors are stimulation condition (control no-TMS, real TMS, and sham TMS) and
practice order (blocked and random). Since TMS pulses could have influenced motor
behavior through elements other than cortical perturbation (e.g., the coil-discharged
noise and sensation on the head), three stimulation conditions were included to
systematically rule out alternative interpretations of the TMS effects on motor
performance and learning.
This chapter begins by summarizing the main results with reference to each a
priori hypothesis. Then I will discuss how this dissertation draws upon and
contributes to the literature in the behavioral and cognitive neurosciences. Alternative
interpretations, limitations, and potential clinical implications that are important for
future hypothesis generation, are also discussed.
Summary of Main Findings
Four hypotheses corresponding to four specific aims were proposed.
Hypothesis 1: The CI effect would generalize to the learning of fast discrete goal-
directed arm movements. Participants who practice under a random order condition
would demonstrate better retention performance than those who practice under a
64
blocked order condition. Hypothesis 2: the cortical motor system (CM) is engaged to
a greater extent when participants practice in a higher interference condition (random
order condition) than for a lower interference condition (blocked order condition). If
so, TMS perturbation centered over CM should selectively disrupt motor learning
under random practice compared to a no perturbation condition. Hypothesis 3: CM is
a putative neural substrate for processing task-relevant timing and/or amplitude.
Perturbation over CM should interfere with learning of these kinematic details
compared with a no perturbation condition. Hypothesis 4: If the elaborative
processing hypothesis accounts for the CI effect, then the advantage of random
practice should be diminished by TMS, applied during the inter-trial interval. If the
forgetting reconstruction hypothesis accounts for the CI effect, then blocked practice
with TMS should become more like random practice when TMS is applied during the
inter-trial interval.
The first hypothesis was supported (Chapter 2) by showing that random order
practice was more beneficial for motor skill learning compared with that for blocked
order practice as revealed by delayed retention performance. More importantly, in
addition to confirming previous findings that CM is a putative center for motor
learning, we demonstrated that CM selectively engaged in processing task-relevant
information when learners practiced under a random order condition, and thus
supporting the second hypothesis (Grafton et al., 1992, Kawashima et al., 1994, Elbert
et al., 1995, Nudo et al., 1996, Classen et al., 1998, Cohen et al., 1998, Karni et al.,
1998, Kleim et al., 1998, van Mier et al., 1998, Classen et al., 1999, Hess et al., 1999,
65
Hund-Georgiadis and von Cramon, 1999, Liepert et al., 1999, Frost et al., 2000, Plautz
et al., 2000, Butefisch et al., 2004, Pleger et al., 2004, van Mier et al., 2004, Ziemann
et al., 2004, Kleim et al., 2006).
The third hypothesis was only partially supported (Chapter 3). We
demonstrated that TMS perturbation of CM interfered with learning of movement
timing, but not amplitude. This pattern of results suggests that CM is a neural
substrate for processing temporal information during motor skill learning. Another
noteworthy finding discussed in Chapter 3 is that this form of cortical perturbation did
not interfere significantly with spatial parameter learning. Previous studies have
identified a role for CM in planning, adjusting, and encoding task-specific spatial
information (Crowe et al., 2004, Diedrichsen et al., 2005, Liu et al., 2005, Monfils et
al., 2005, Chouinard and Paus, 2006, Coxon et al., 2006, Lee and van Donkelaar,
2006, Naselaris et al., 2006a). We speculate that this processing may be modulated by
the strategy participants adopt to improve task performance. For example, in this and
previous studies conducted in our laboratory, the best strategy for participants to
reduce error in the lever arm tasks is to improve movement timing (Lin et al., 2007a).
Therefore, it is conceivable that CM is less tuned to processing movement amplitude
since the behavioral preference for this task is to optimize movement timing.
Results from the third study (Chapter 4) support the elaborative processing hypothesis
by showing that if the TMS pulse is timed during the inter-trial interval to likely
disrupt task-elaboration processes, the learning benefit of random practice is
attenuated. The forgetting reconstruction view was not supported given that the TMS
66
perturbation during practice did not benefit motor learning for participants who
practiced in the blocked order condition compared with those in the no-TMS blocked
order condition. Further opposition to the forgetting reconstruction hypothesis are the
data of the acquisition phase. The forgetting reconstruction hypothesis predicts that
performance should deteriorate if forgetting of a previously constructed action plan
was induced (Lee and Magill, 1983). However, this is not the case for the Blocked
practice groups under TMS perturbation conditions. Instead, acquisition performance
was mildly enhanced under TMS conditions during blocked practice, suggesting that
the perturbation was working through mechanisms other than those associated with
forgetting-induced reconstruction. Post-experiment interviews in earlier work by
Shea and Zimny (1983), indicated that participants in the blocked condition tended to
run the movements off without much thought, more or less “automatically” (Shea and
Zimny, 1983), we therefore speculate that the additional TMS perturbation during
blocked practice may have necessitated a reallocation of attentional resources to the
motor tasks, resulting in better performance during acquisition and delayed retention.
This interpretation is suggested by the more apparent performance improvement
during the early acquisition phase and the more accurate performance during the
delayed retention phase of the Blocked-TMS group compared to the Blocked-control
group (i.e., trend without significance).
The results shed light on the nature of motor memory consolidation
Though this dissertation was not designed to directly probe the nature of motor
memory consolidation, we found that the perturbation effect from TMS to motor
67
learning was not robustly apparent until the delayed retention phase (Chapters 2 and
3). These findings underscore the importance of the Learning Performance
Distinction fundamental to learning and memory research. In particular, these findings
clearly demonstrate that not all changes during acquisition represent the more
permanent changes of behavior (the definition of “learning”) (Cahill et al., 2001,
Schmidt and Lee, 2005). We suggest that the TMS perturbation applied during the
inter-trial interval inhibits memory consolidation of task-relevant information.
Perturbation centered on and around the motor cortex during practice led to a poorer
motor memory formation particularly when the practice condition required greater CM
engagement for inter-task comparisons and parameter specification during random
order condition. The magnetic pulse delivered at each inter-trial interval created a
transient “lesion” and appears to have interfered with the consolidation of task-
relevant information. Taken together, we propose that CM is functionally important
for consolidation of task-relevant information; an interpretation that is consistent with
previous studies in which the cortical motor system has been shown to be a neural
substrate for motor memory formation (Shadmehr and Brashers-Krug, 1997,
Muellbacher et al., 2002, Robertson, 2004, Robertson et al., 2005, Krakauer and
Shadmehr, 2006). The fact that TMS did not interfere with memory consolidation for
participants who practiced in the blocked order condition suggests that CM is likely
not the final storage area for motor memory.
Further, this dissertation is an inter-disciplinary project that draws from the
behavioral science (motor learning) and cognitive neuroscience. From the motor
68
learning perspective, the dissertation started with a research question about a motor
learning phenomenon (contextual interference effect) and extended the scope by
exploring the neural substrates implementing this phenomenon (Chapter 2). This is
the first study attempting to replicate the contextual interference effect by introducing
a neural “interference” at the central level (the corticomotor system) when subjects
engage in motor practice (Chapter 4). In the following section, I will discuss how
findings of this dissertation shed light on two different but complimentary conceptual
frameworks for skill learning, the Challenge Point Framework and the Control-Based
Learning Theory. This discussion is followed by a section that describes how these
results provide insight into one of the fundamental principles of experience-dependent
neural plasticity.
Our results provide support for the recently proposed Challenge Point
Framework (Guadagnoli and Lee, 2004). The Challenge Point Framework suggests
that there is an ideal level of functional task difficulty which determines how much
information will be available for motor learning. Functional task difficulty is defined
as the interaction of nominal task difficulty, learner characteristics, and specific
practice conditions. This Framework predicts that when the learner’s characteristics
and nominal task difficulty are controlled, optimal learning likely occurs when the
ideal “amount” of information processing is evoked through specific conditions of
practice. An inference from this prediction, in the circumstance of the CI effect, is that
if greater levels of interference during motor practice can enhance learning, there is an
ideal level of interference that should effect optimal motor learning.
69
It appears that these findings support the prediction of the Challenge Point
Framework. TMS perturbation during blocked order practice mildly enhanced recall
accuracy compared to that in the no-TMS blocked condition. On the other hand, TMS
deteriorated recall accuracy for learners who practiced in the random order condition.
This suggests that increased interference during practice is beneficial to motor learning
only if the interference does not exceed the “challenge point”. For an initially
“challenging” practice condition, such as that induced through the random condition,
increasing the interference with TMS may have exceeded the “challenge point” and
resulted in poorer recall performance.
Our findings further support the prediction of the Control-Based Learning
Theory (Willingham, 1998) which suggests that motor learning grows directly out of
motor control processes. This theory predicts that in addition to its known role as a
motor execution center, CM is also a putative center for motor learning. Motor
learning emerges as CM becomes more efficient in processing task-relevant
information (Tanji, 1996, Hund-Georgiadis and von Cramon, 1999, Lee et al., 1999,
Meister et al., 2005). Results of Chapters 2 and 3 support this theory by
demonstrating that CM has an essential role in motor skill learning, likely via
processing and consolidation of task-relevant information such as movement timing.
Our findings demonstrate a fundamental principle of the central nervous
system (CNS): plasticity can be experience-dependent (Kleim and Jones). In
particular, this principle suggests that the nature of the training experience dictates the
nature of the plasticity. In Chapter 2, we demonstrated that practice context
70
characterized the training experience. In the no-TMS condition, positive plasticity
(motor learning) was induced under random practice conditions. However, when
TMS was used to perturb the encoding of that training experience, negative plasticity
resulted (poorer learning). Our experimental design with TMS informs us about the
contribution of the cortical motor system in experience-dependent learning and
plasticity. These findings provide a better understanding of how training context and
experience may interact with the neural implementation driven by the cortical motor
system, and further support the use of evidence-based training protocols to drive brain
reorganization and motor learning in neurorehabilitation.
Clinical Implications
This dissertation work has important implications for understanding the
neuromotor processes by which the learning of fast discrete goal-directed arm
movements are accomplished under different practice conditions. We confirmed the
benefits of random order practice in facilitating motor skill learning. Our TMS study
design provided a mechanistic way of combining task training order with cortical-
location specific therapies where the brain is directly stimulated. Existing therapies
that are still in the development phase include repetitive TMS (rTMS), transcranial
direct current stimulation (tDCS), or cortical electrical stimulation by a stimulator
directly implanted in the cortex. For example, findings of Chapters 2 and 3 together
suggest that a suprathreshold level of single-pulse magnetic stimulation appropriately
timed during practice can influence motor learning. This finding provides possible
beneficial effects of such adjunctive therapies for motor rehabilitation. What remains
71
to be determined and provides for the future direction of this work, is to identify: 1)
the stimulation parameters which can be safely and effectively applied in conjunction
with various motor training protocols; and 2) the stimulation parameters that are safe
and effective for specific disease populations.
Limitations, Alternative Interpretations, Future Directions
Timing of TMS pulses
A main finding in this dissertation is that single-pulse modulation precisely
timed within the inter-trial interval during the practice phase produces significant
effects on delayed recall performance for random but not blocked practice conditions.
This conclusion is limited because we did not investigate the effect of TMS
modulation during other time windows during motor practice. However, the apparent
dissociation in CM activity between blocked and random practice conditions at this
particular time window strengthens the validity of our design and results. Future
studies designed to systematically probe the inter-trial interval or with other
neurophysiologic methods such as magnetoencephalography (MEG) or
electroencephalography (EEG) may be helpful to obtain greater resolution on the
timing for putative processes of motor learning.
Choice of stimulating coil
Another factor which may alter our interpretation is that the magnetic
stimulation could have spread to neighboring neural substrates such as the dorsal
premotor and somatosensory cortices. Though we tried to ensure the specificity of
TMS pulses over the motor cortex by our method of 1) placing one edge of the coil on
72
the subject’s scalp closest to the motor represetntaion of contralateral biceps (Kamen,
2004), and 2) recording electromyography of each evoked muscle response throughout
data collection, we acknowledge the likelihood that the circular coil produces current
flow that would include the dorsal premotor cortex (dPMC) and adjacent
somatosensory cortex (S1). The dPMC has been implicated in motor learning through
visuomotor associations and through selecting from a group of motor programs (Lee
and van Donkelaar, 2006, Praeg et al., 2006), operations that were likely involved here.
S1 shares significant overlap in the corticomotor system with M1 and likely processes
somatosensory feedback from the limb with each movement (Dancause, 2006, Floyer-
Lea et al., 2006). Although the choice of the circular-shaped coil limits specificity of
the cortical regions perturbed, it is clear that our study design highlights a cortical-
specific influence in primary motor or motor planning cortices that is selectively
present during random, but not blocked, practice of motor skills.
Factors other than cortical perturbation
Our interpretation does not rule out the possibility that the function of CM
could be perturbed perhaps in ways other than through direct cortical stimulation. For
example, the peripheral muscle twitches induced by suprathreshold TMS could
potentially "perturb" the cortical motor system through somatosensory feedback. It is
also hard to separate the role of motor practice from that of the attentional load in
eliciting CM plasticity. It is possible, however, that both factors are important and
contribute to this effect.
73
Time course of parameter specifications
Though this dissertation did not address the question about the time course at
which the specification of kinematic details emerges during skill acquisition, the
results of a companion project
147
suggest the possibility that the learner focuses on the
temporal parameter during the early acquisition phase, at least the first 144 trials, and
then begins to scale the spatial parameter. Further studies are necessary to investigate
the evolution of the spatial and temporal parameterization in the development of motor
skills. It would be interesting to know how different motor parameters evolve and
merge in the process of motor skill learning and what could be done during practice to
enhance that natural evolution for more efficient motor learning.
Task issue
As discussed in Chapter 2, the nature of motor tasks, as well as the stimulation
parameters, likely influence the way in which TMS modulates motor cortex activity
during motor practice and subsequent memory formation (Muellbacher et al., 2002)
,
(Baraduc et al., 2004). Using a single-pulse disruptive paradigm, we selectively
disrupted motor learning for subjects who practiced under a random condition, but not
blocked condition. This is in contrast to recent results by Muellbacher et al.
(Muellbacher et al., 2002) and Baraduc et al. (Baraduc et al., 2004), that used
repetitive TMS (rTMS) over primary motor cortex. These investigators arranged
practice in a blocked order and were able to disrupt early memory consolidation of
ballistic finger movements. We speculate three possible reasons that could give rise to
differences in findings between this and our work. First, the neural bases of learning
74
coordinated tasks is known to be substantially differs from those involved in learning
simple ballistic tasks, particularly at it relates to the primary motor cortex(Baraduc et
al., 2004). Second, our single-pulse TMS disruption during blocked practice may not
be as disruptive as the rTMS applied in the Muellbacher et al. and Baraduc et al.
experiments. The distributed, single-pulse magnetic perturbation used herein may not
have accumulated as much disruptive effect as rTMS to the motor engrams of the
trained tasks. Future investigation of the post-TMS excitability changes induced by
single-pulse TMS and their correlation with task performance indices would help shed
light on this issue. Third, the motor engrams for the coordinated arm task could be
distributed across several cerebral areas, such as the supplementary motor area (Wise
et al., 1998), premotor cortex (Mitz et al., 1991, Shen and Alexander, 1997, Wise et al.,
1998), parietal cortex (Eskandar and Assad, 1999), and cerebellum (Ebner and Fu,
1997, Liu et al., 2003). Whether these neural substrates directly participate in
encoding or consolidating of motor memory remains to be determined. Alternatively,
it has been proposed that learning the dynamics of a coordinated task leads to the
formation of an internal model (Kawato and Wolpert, 1998), which could be stored in
the cerebellum (Blakemore et al., 1999, Imamizu et al., 2000, Liu et al., 2003). All
these possibilities lead to novel research questions and hypotheses worthy of future
study.
Lastly, the laboratory tasks we used are analogs to real world skills, bringing
into question the generalizability of task learning in a real-world environment with its
meaningful motor skills. Future work which applies more “real-life” and motivating
75
motor tasks and practice conditions is necessary to test the generalizability of our
findings. In spite of this limitation, this dissertation work provides preliminary data
and new insights into the nature of the contextual interference effect in motor learning.
It also provides a beginning foundation for the translation of motor learning principles
for advancing and developing innovative therapeutic approaches in neurorehabilitation.
76
REFERENCES
Albaret, J. M. and Thon, B., 1998. Differential effects of task complexity on
contextual interference in a drawing task. Acta Psychol (Amst). 100, 9-24.
Amassian, V. E., Cracco, R. Q., Maccabee, P. J., Cracco, J. B., Rudell, A. P. and
Eberle, L., 1998. Transcranial magnetic stimulation in study of the visual
pathway. J Clin Neurophysiol. 15, 288-304.
Aoki, T., Tsuda, H., Takasawa, M., Osaki, Y., Oku, N., Hatazawa, J. and Kinoshita, H.,
2005. The effect of tapping finger and mode differences on cortical and
subcortical activities: a PET study. Exp Brain Res. 160, 375-383.
Baraduc, P., Lang, N., Rothwell, J. C. and Wolpert, D. M., 2004. Consolidation of
dynamic motor learning is not disrupted by rTMS of primary motor cortex.
Curr Biol. 14, 252-256.
Baranyi, A. and Szente, M. B., 1987. Long-lasting potentiation of synaptic
transmission requires postsynaptic modifications in the neocortex. Brain Res.
423, 378-384.
Baranyi, A., Szente, M. B. and Woody, C. D., 1991. Properties of associative long-
lasting potentiation induced by cellular conditioning in the motor cortex of
conscious cats. Neuroscience. 42, 321-334.
Battig, W. F., 1979. The flexibility of human memory. In: Cermak, L. S. and Craik, F.
I. M. (Eds.), Level of Processing in Human Memor. Erlbaum, Hillsdale, NJ, pp.
23-44.
Ben-Shaul, Y., Drori, R., Asher, I., Stark, E., Nadasdy, Z. and Abeles, M., 2004.
Neuronal activity in motor cortical areas reflects the sequential context of
movement. J Neurophysiol. 91, 1748-1762.
Berardelli, A., Inghilleri, M., Polidori, L., Priori, A., Mercuri, B. and Manfredi, M.,
1994. Effects of transcranial magnetic stimulation on single and sequential arm
movements. Exp Brain Res. 98, 501-506.
Bestmann, S., Siebner, H. R., Modugno, N., Amassian, V. E. and Rothwell, J. C.,
2004. Inhibitory interactions between pairs of subthreshold conditioning
stimuli in the human motor cortex. Clin Neurophysiol. 115, 755-764.
Blakemore, S. J., Wolpert, D. M. and Frith, C. D., 1999. The cerebellum contributes to
somatosensory cortical activity during self-produced tactile stimulation.
Neuroimage. 10, 448-459.
77
Boroojerdi, B., Phipps, M., Kopylev, L., Wharton, C. M., Cohen, L. G. and Grafman,
J., 2001. Enhancing analogic reasoning with rTMS over the left prefrontal
cortex. Neurology. 56, 526-528.
Bortoli, L., Robazza, C., Durigon, V. and Carra, C., 1992. Effects of contextual
interference on learning technical sports skills. Percept Mot Skills. 75, 555-562.
Boyce, B. A. and Del Rey, P., 1990. Designing applied research in a naturalistic
setting using a contextual interference paradigm. J Human Movement Studies.
18, 189-200.
Braun, C., Haug, M., Wiech, K., Birbaumer, N., Elbert, T. and Roberts, L. E., 2002.
Functional organization of primary somatosensory cortex depends on the focus
of attention. Neuroimage. 17, 1451-1458.
Butefisch, C. M., Khurana, V., Kopylev, L. and Cohen, L. G., 2004. Enhancing
encoding of a motor memory in the primary motor cortex by cortical
stimulation. J Neurophysiol. 91, 2110-2116.
Cahill, L., McGaugh, J. L. and Weinberger, N. M., 2001. The neurobiology of
learning and memory: some reminders to remember. Trends Neurosci. 24, 578-
581.
Carnahan, H., Van Eerd, D. L. and Allard, F., 1990. A noted on the relationship
between task requirements and the contextual interference effect. J Mot Behav.
22, 159-169.
Chouinard, P. A. and Paus, T., 2006. The primary motor and premotor areas of the
human cerebral cortex. Neuroscientist. 12, 143-152.
Chung, S. C., Min, B. C., Kim, C. J. and Cho, Z. H., 2000. Total activation change of
visual and motor area due to various disturbances. J Physiol Anthropol Appl
Human Sci. 19, 93-100.
Classen, J., Liepert, J., Hallett, M. and Cohen, L., 1999. Plasticity of movement
representation in the human motor cortex. Electroencephalogr Clin
Neurophysiol Suppl. 51, 162-173.
Classen, J., Liepert, J., Wise, S. P., Hallett, M. and Cohen, L. G., 1998. Rapid
plasticity of human cortical movement representation induced by practice. J
Neurophysiol. 79, 1117-1123.
Cohen, L. G., Ziemann, U., Chen, R., Classen, J., Hallett, M., Gerloff, C. and
Butefisch, C., 1998. Studies of neuroplasticity with transcranial magnetic
stimulation. J Clin Neurophysiol. 15, 305-324.
78
Coxon, J. P., Stinear, C. M. and Byblow, W. D., 2006. Intracortical inhibition during
volitional inhibition of prepared action. J Neurophysiol. 95, 3371-3383.
Crowe, D. A., Chafee, M. V., Averbeck, B. B. and Georgopoulos, A. P., 2004.
Participation of primary motor cortical neurons in a distributed network during
maze solution: representation of spatial parameters and time-course
comparison with parietal area 7a. Exp Brain Res. 158, 28-34.
Dancause, N., 2006. Neurophysiological and anatomical plasticity in the adult
sensorimotor cortex. Rev Neurosci. 17, 561-580.
Del Rey, P., 1982. Effects of contextual interference on the memory of older females
differing in levels of physical activity. Percept Mot Skills. 55, 171-180.
Del Rey, P., 1989. Training and contextual interference effects on memory and
transfer. Res Q Exerc Sport. 60, 342-347.
Del Rey, P., Liu, X. and Simpson, K. J., 1994. Does retroactive inhibition influence
contextual interference effects? Res Q Exerc Sport. 65, 120-126.
Del Rey, P., Wughalter, E. H. and Whitehurst, M., 1982. The effects of contextual
interference on females with varied experience in open sport skills. Res Q
Exerc Sport. 53, 108-115.
Demiralp, T., Karamursel, S., Karakullukcu, Y. E. and Gokhan, N., 1990. Movement-
related cortical potentials: their relationship to the laterality, complexity and
learning of a movement. Int J Neurosci. 51, 153-162.
Desmurget, M., Epstein, C. M., Turner, R. S., Prablanc, C., Alexander, G. E. and
Grafton, S. T., 1999. Role of the posterior parietal cortex in updating reaching
movements to a visual target. Nat Neurosci. 2, 563-567.
Diedrichsen, J., Hashambhoy, Y., Rane, T. and Shadmehr, R., 2005. Neural correlates
of reach errors. J Neurosci. 25, 9919-9931.
Ebner, T. J. and Fu, Q., 1997. What features of visually guided arm movements are
encoded in the simple spike discharge of cerebellar Purkinje cells? Prog Brain
Res. 114, 431-447.
Elbert, T., Pantev, C., Wienbruch, C., Rockstroh, B. and Taub, E., 1995. Increased
cortical representation of the fingers of the left hand in string players. Science.
270, 305-307.
Eskandar, E. N. and Assad, J. A., 1999. Dissociation of visual, motor and predictive
signals in parietal cortex during visual guidance. Nat Neurosci. 2, 88-93.
79
Flitman, S. S., Grafman, J., Wassermann, E. M., Cooper, V., O'Grady, J., Pascual-
Leone, A. and Hallett, M., 1998a. Linguistic processing during repetitive
transcranial magnetic stimulation. Neurology. 50, 175-181.
Flitman, S. S., Grafmann, J., Wassermann, E. M., Cooper, V., O'Grady, J., Pascual-
Leone, A. and Hallett, M., 1998b. Linguistic processing during repetitive
transcranial magnetic stimulation.
. Neurology. 50, 175-181.
Floyer-Lea, A., Wylezinska, M., Kincses, T. and Matthews, P. M., 2006. Rapid
modulation of GABA concentration in human sensorimotor cortex during
motor learning. J Neurophysiol. 95, 1639-1644.
Frost, S. B., Milliken, G. W., Plautz, E. J., Masterton, R. B. and Nudo, R. J., 2000.
Somatosensory and motor representations in cerebral cortex of a primitive
mammal (Monodelphis domestica): a window into the early evolution of
sensorimotor cortex. J Comp Neurol. 421, 29-51.
Fu, Y. X., Djupsund, K., Gao, H., Hayden, B., Shen, K. and Dan, Y., 2002. Temporal
specificity in the cortical plasticity of visual space representation. Science. 296,
1999-2003.
Gabriele, T. E., Hall, G. R. and Buckolz, E. E., 1987. Practice schedule effects on the
acquisition and retention of motor skill. Human Movement Science. 6, 1-16.
Giuffrida, C. G., Shea, J. B. and Fairbrother, J. T., 2002. Differential transfer benefits
of increased practice for constant, blocked, and serial practice schedules. J Mot
Behav. 34, 353-365.
Goode, S. L., Meeuwsen, H. J. and Magill, R. A., 1998. Benefits of providing
cognitive performance strategies to novice performers learning a complex
motor skill. Percept Mot Skills. 86, 976-978.
Goodwin, J. E. and Meeuwsen, H. J., 1996. Investigation of the contextual
interference effect in the manipulation of the motor parameter of over-all force.
Percept Mot Skills. 83, 735-743.
Gorbet, D. J., Staines, W. R. and Sergio, L. E., 2004. Brain mechanisms for preparing
increasingly complex sensory to motor transformations. Neuroimage. 23,
1100-1111.
Grafton, S. T., Mazziotta, J. C., Presty, S., Friston, K. J., Frackowiak, R. S. and Phelps,
M. E., 1992. Functional anatomy of human procedural learning determined
with regional cerebral blood flow and PET. J Neurosci. 12, 2542-2548.
80
Gregori, B., Curra, A., Dinapoli, L., Bologna, M., Accornero, N. and Berardelli, A.,
2005. The timing and intensity of transcranial magnetic stimulation, and the
scalp site stimulated, as variables influencing motor sequence performance in
healthy subjects. Exp Brain Res. 166, 43-55.
Guadagnoli, M. A. and Lee, T. D., 2004. Challenge point: a framework for
conceptualizing the effects of various practice conditions in motor learning. J
Mot Behav. 36, 212-224.
Hall, K. G., Domingues, D. A. and Cavazos, R., 1994. Contextual interference effects
with skilled baseball players. Percept Mot Skills. 78, 835-841.
Hall, K. G. and Magill, R. A., 1995. Variability of Practice and Contextual
Interference in Motor Skill Learning. J Mot Behav. 27, 299-309.
Hallett, M., 2000. Transcranial magnetic stimulation and the human brain. Nature. 406,
147-150.
Hatfield, B. D., Haufler, A. J., Hung, T. M. and Spalding, T. W., 2004.
Electroencephalographic studies of skilled psychomotor performance. J Clin
Neurophysiol. 21, 144-156.
Hebb, D. O., 1949. The Organization of Behaviour. . John Wiley & Sons, New York.
Hess, A., Kunesch, E., Classen, J., Hoeppner, J., Stefan, K. and Benecke, R., 1999.
Task-dependent modulation of inhibitory actions within the primary motor
cortex. Exp Brain Res. 124, 321-330.
Hund-Georgiadis, M. and von Cramon, D. Y., 1999. Motor-learning-related changes
in piano players and non-musicians revealed by functional magnetic-resonance
signals. Exp Brain Res. 125, 417-425.
Imamizu, H., Miyauchi, S., Tamada, T., Sasaki, Y., Takino, R., Putz, B., Yoshioka, T.
and Kawato, M., 2000. Human cerebellar activity reflecting an acquired
internal model of a new tool. Nature. 403, 192-195.
Jelsma, O. and Pieters, J. M., 1989. Practice schedule and cognitive style interaction in
learning a maze task. Applied Cognitive Psychology 3, 73-83.
Jelsma, O. and Van Merrienboer, J., 1989. Contextual interference: interactions with
reflection-impulsivity Percept Mot Skills. 68, 1055-1964.
Kahneman, D., 1973. Attention and effort. Prentice-Hall, Englewood Cliffs, NJ.
Kamen, G., 2004. Reliability of motor-evoked potentials during resting and active
contraction conditions. Med Sci Sports Exerc. 36, 1574-1579.
81
Kaneko, K., Kawai, S., Fuchigami, Y., Shiraishi, G. and Ito, T., 1996. Effect of
stimulus intensity and voluntary contraction on corticospinal potentials
following transcranial magnetic stimulation. J Neurol Sci. 139, 131-136.
Karni, A., Meyer, G., Rey-Hipolito, C., Jezzard, P., Adams, M. M., Turner, R. and
Ungerleider, L. G., 1998. The acquisition of skilled motor performance: fast
and slow experience-driven changes in primary motor cortex. Proc Natl Acad
Sci U S A. 95, 861-868.
Kawashima, R., Roland, P. E. and O'Sullivan, B. T., 1994. Fields in human motor
areas involved in preparation for reaching, actual reaching, and visuomotor
learning: a positron emission tomography study. J Neurosci. 14, 3462-3474.
Kawato, M. and Wolpert, D., 1998. Internal models for motor control. Novartis Found
Symp. 218, 291-304; discussion 304-297.
Kleim, J. A., Barbay, S., Cooper, N. R., Hogg, T. M., Reidel, C. N., Remple, M. S.
and Nudo, R. J., 2002. Motor learning-dependent synaptogenesis is localized to
functionally reorganized motor cortex. Neurobiol Learn Mem. 77, 63-77.
Kleim, J. A., Barbay, S. and Nudo, R. J., 1998. Functional reorganization of the rat
motor cortex following motor skill learning. J Neurophysiol. 80, 3321-3325.
Kleim, J. A., Chan, S., Pringle, E., Schallert, K., Procaccio, V., Jimenez, R. and
Cramer, S. C., 2006. BDNF val66met polymorphism is associated with
modified experience-dependent plasticity in human motor cortex. Nat Neurosci.
9, 735-737.
Kleim, J. A. and Jones, T. A., Principles of Experience-Dependent Neural Plasticity:
Implications for
Rehabilitation after Brain Damage. Journal of Speech Hearing and Language Research.
(in press).
Kleim, J. A. and Jones, T. A., 2007a. Principles of Experience-Dependent Neural
Plasticity: Implications for
Rehabilitation after Brain Damage), Gainesville, FL.
Kleim, J. A. and Jones, T. A., 2007b. Principles of Experience-Dependent Neural
Plasticity: Implications for
Rehabilitation after Brain Damage. Journal of Speech Hearing and Language Research.
(in press).
Kleim, J. A., Jones, T. A. and Schallert, T., 2003. Motor enrichment and the induction
of plasticity before or after brain injury. Neurochem Res. 28, 1757-1769.
82
Krakauer, J. W. and Shadmehr, R., 2006. Consolidation of motor memory. Trends
Neurosci. 29, 58-64.
Kuriki, S., Kanda, S. and Hirata, Y., 2006. Effects of musical experience on different
components of MEG responses elicited by sequential piano-tones and chords. J
Neurosci. 26, 4046-4053.
Lee, D., Port, N. L., Kruse, W. and Georgopoulos, A. P., 2001. Neuronal clusters in
the primate motor cortex during interception of moving targets. J Cogn
Neurosci. 13, 319-331.
Lee, J. H. and van Donkelaar, P., 2006. The human dorsal premotor cortex generates
on-line error corrections during sensorimotor adaptation. J Neurosci. 26, 3330-
3334.
Lee, K. M., Chang, K. H. and Roh, J. K., 1999. Subregions within the supplementary
motor area activated at different stages of movement preparation and execution.
Neuroimage. 9, 117-123.
Lee, T. D., 1988. Transfer-appropriate processing: a framework for conceptualizing
practice effects in motor learning In: Meijer, O. G. and Roth, K. (Eds.),
Complex Movement Behaviour: The motor-action controversy. Elsevier
Science Publishers B.V., North-Holland, pp. 201-215.
Lee, T. D. and Magill, R. A., 1983. The locus of contextual interference in motor-skill
acquisition. J Exp Psychol: Human Learning and Memory. 9, 730-746.
Lee, T. D. and Magill, R. A., 1985. Can forgetting facilitate skill acquisition? In:
Goodman, D. et al. (Eds.), Differing Perspectives in Motor Learning, Memory,
and Control. Elsevier Science Publishers B.V., North-Holland, pp. 3-21.
Lee, T. D., Magill, R. A. and Weeks, D. J., 1985. Influence of practice schedule on
testing schema theory predictions in adults. J Mot Behav. 17, 283-299.
Lee, T. D., Wishart, L. R., Cunningham, S. and Carnahan, H., 1997. Modeled timing
information during random practice eliminates the contextual interference
effect. Res Q Exerc Sport. 68, 100-105.
Lee, T. D., Wulf, G. and Schmidt, R. A., 1992. Contextual interference in motor skill
learning: dissociated effects due to the nature of task variations. Quartly
Journal Experimental Psychology. 44, 627-644.
Lewis, P. A., Wing, A. M., Pope, P. A., Praamstra, P. and Miall, R. C., 2004. Brain
activity correlates differentially with increasing temporal complexity of
rhythms during initialisation, synchronisation, and continuation phases of
paced finger tapping. Neuropsychologia. 42, 1301-1312.
83
Li, Y. and Wright, D. L., 2000. An assessment of the attention demands during
random- and blocked-practice schedules. Q J Exp Psychol A. 53, 591-606.
Liepert, J., Terborg, C. and Weiller, C., 1999. Motor plasticity induced by
synchronized thumb and foot movements. Exp Brain Res. 125, 435-439.
Limons, E. and Shea, J. B., 1988. Deficient processing in learning and performance. In:
Colley, A. M. and Beech, J. R. (Eds.), Cognition and Action in Skilled Behavio.
Elsevier Science Publisher, North-Holland, pp. 333-347.
Lin, C., Winstein, C. J., Wu, A. D., Lee, L. and Gordon, J., 2007a. Skill development
of fast sequential arm movements: the contribution of motor parameters and
practice orders), Los Angeles, CA.
Lin, C., Wu, A. D., Fisher, B. E. and Winstein, C. J., 2007b. Task Practice Order and
the Role of Human Motor Cortex in Motor Skill Learning: A Transcranial
Magnetic Stimulation (TMS) Investigation. J Neurophysiol. (under review).
Liu, X., Robertson, E. and Miall, R. C., 2003. Neuronal activity related to the visual
representation of arm movements in the lateral cerebellar cortex. J
Neurophysiol. 89, 1223-1237.
Liu, Y., Denton, J. M. and Nelson, R. J., 2005. Neuronal activity in primary motor
cortex differs when monkeys perform somatosensory and visually guided wrist
movements. Exp Brain Res. 167, 571-586.
Macar, F., Coull, J. and Vidal, F., 2006. The supplementary motor area in motor and
perceptual time processing: fMRI studies. Cogn Process. 7, 89-94.
Magill, R. A. and Hall, K. G., 1990. A review of the contextual interference effect in
motor skill acquisition. . Human Movement Science. 9, 241-289.
Martin, T., Houck, J. M., Bish, J. P., Kicic, D., Woodruff, C. C., Moses, S. N., Lee, D.
C. and Tesche, C. D., 2006. MEG reveals different contributions of
somatomotor cortex and cerebellum to simple reaction time after temporally
structured cues. Hum Brain Mapp. 27, 552-561.
Meira, C. M., Jr. and Tani, G., 2001. The contextual interference effect in acquisition
of dart-throwing skill tested on a transfer test with extended trials. Percept Mot
Skills. 92, 910-918.
Meister, I., Krings, T., Foltys, H., Boroojerdi, B., Muller, M., Topper, R. and Thron,
A., 2005. Effects of long-term practice and task complexity in musicians and
nonmusicians performing simple and complex motor tasks: implications for
cortical motor organization. Hum Brain Mapp. 25, 345-352.
84
Mitz, A. R., Godschalk, M. and Wise, S. P., 1991. Learning-dependent neuronal
activity in the premotor cortex: activity during the acquisition of conditional
motor associations. J Neurosci. 11, 1855-1872.
Monfils, M. H., Plautz, E. J. and Kleim, J. A., 2005. In search of the motor engram:
motor map plasticity as a mechanism for encoding motor experience.
Neuroscientist. 11, 471-483.
Muellbacher, W., Ziemann, U., Wissel, J., Dang, N., Kofler, M., Facchini, S.,
Boroojerdi, B., Poewe, W. and Hallett, M., 2002. Early consolidation in human
primary motor cortex. Nature. 415, 640-644.
Naselaris, T., Merchant, H., Amirikian, B. and Georgopoulos, A. P., 2006a. Large-
scale organization of preferred directions in the motor cortex. I. Motor cortical
hyperacuity for forward reaching. J Neurophysiol. 96, 3231-3236.
Naselaris, T., Merchant, H., Amirikian, B. and Georgopoulos, A. P., 2006b. Large-
scale organization of preferred directions in the motor cortex. II. Analysis of
local distributions. J Neurophysiol. 96, 3237-3247.
Nudo, R. J., Milliken, G. W., Jenkins, W. M. and Merzenich, M. M., 1996. Use-
dependent alterations of movement representations in primary motor cortex of
adult squirrel monkeys. J Neurosci. 16, 785-807.
Pascual-Leone, A., Brasil-Neto, J. P., Valls-Sole, J., Cohen, L. G. and Hallett, M.,
1992a. Simple reaction time to focal transcranial magnetic stimulation.
Comparison with reaction time to acoustic, visual and somatosensory stimuli.
Brain. 115 Pt 1, 109-122.
Pascual-Leone, A., Houser, C. M., Grafman, J. and Hallett, M., 1992b. Reaction time
and transcranial magnetic stimulation. Lancet. 339, 1420.
Perez, C. R., Meira, C. M., Jr. and Tani, G., 2005. Does the contextual interference
effect last over extended transfer trials? Percept Mot Skills. 100, 58-60.
Plautz, E. J., Milliken, G. W. and Nudo, R. J., 2000. Effects of repetitive motor
training on movement representations in adult squirrel monkeys: role of use
versus learning. Neurobiol Learn Mem. 74, 27-55.
Pleger, B., Schwenkreis, P., Grunberg, C., Malin, J. P. and Tegenthoff, M., 2004.
Fluoxetine facilitates use-dependent excitability of human primary motor
cortex. Clin Neurophysiol. 115, 2157-2163.
Praeg, E., Esslen, M., Lutz, K. and Jancke, L., 2006. Neuronal modifications during
visuomotor association learning assessed by electric brain tomography. Brain
Topogr. 19, 61-75.
85
Robertson, E. M., 2004. Skill learning: putting procedural consolidation in context.
Curr Biol. 14, R1061-1063.
Robertson, E. M., Press, D. Z. and Pascual-Leone, A., 2005. Off-line learning and the
primary motor cortex. J Neurosci. 25, 6372-6378.
Rossini, P. M., Martino, G., Narici, L., Pasquarelli, A., Peresson, M., Pizzella, V.,
Tecchio, F., Torrioli, G. and Romani, G. L., 1994. Short-term brain 'plasticity'
in humans: transient finger representation changes in sensory cortex
somatotopy following ischemic anesthesia. Brain Res. 642, 169-177.
Schmidt, R. A. and Lee, T. D., 2005. Conditions of practice. In: Schmidt, R. A. and
Lee, T. D. (Eds.), Motor control and learning: A behavior emphasis, vol.11.
Human Kinetics, Champaign, IL, pp. 321-363.
Schubotz, R. I. and von Cramon, D. Y., 2002. A blueprint for target motion: fMRI
reveals perceived sequential complexity to modulate premotor cortex.
Neuroimage. 16, 920-935.
Seidler, R. D., Noll, D. C. and Thiers, G., 2004. Feedforward and feedback processes
in motor control. Neuroimage. 22, 1775-1783.
Sekiya, H., Magill, R. A. and Anderson, D. I., 1996. The contextual interference effect
in parameter modifications of the same generalized motor program. Res Q
Exerc Sport. 67, 59-68.
Sergio, L. E. and Kalaska, J. F., 1998. Changes in the temporal pattern of primary
motor cortex activity in a directional isometric force versus limb movement
task. J Neurophysiol. 80, 1577-1583.
Shadmehr, R. and Brashers-Krug, T., 1997. Functional stages in the formation of
human long-term motor memory. J Neurosci. 17, 409-419.
Shea, C. H., Kohl, R. M. and Indermill, C., 1990. Contextual interference:
contributions of practice. Acta Psychologica. 73, 145-157.
Shea, C. H., Lai, Q., Wright, D. L., Immink, M. and Black, C., 2001. Consistent and
variable practice conditions: effects on relative and absolute timing. J Mot
Behav. 33, 139-152.
Shea, J. B. and Morgan, R., 1979. Contextual interference effects on the acquisition,
retention, and transfer of a motor skill. J Exp Psychol: Human Learning and
Memory. 5, 179-187.
Shea, J. B. and Titzer, R. C., 1993. The influence of reminder trials on contextual
interference effects. J Mot Behav. 25, 264-274.
86
Shea, J. B. and Wright, D. L., 1991. When forgetting benefits motor retention. Res Q
Exerc Sport. 62, 293-301.
Shea, J. B. and Zimny, S. T., 1983. Contextual effects in memory and learning
movement information. . In: Magill, R. A. (Ed.), Memory and Control of
Action Elsevier, Amsterdam, pp. 345-366.
Shea, J. B. and Zimny, S. T., 1988. Knowledge incorporation in motor representation. .
In: Meijer, O. G. and Roth, K. (Eds.), Complex movement behavior: "The"
motor-action controversy. . Elsevier, Amsterdam, pp. 289-314.
Shen, L. and Alexander, G. E., 1997. Preferential representation of instructed target
location versus limb trajectory in dorsal premotor area. J Neurophysiol. 77,
1195-1212.
Sherwood, D. E., 1996. The benefits of random variable practice for spatial accuracy
and error detection in a rapid aiming task. Res Q Exerc Sport. 67, 35-43.
Shewokis, P. A., 1997. Is the contextual interference effect generalizable to computer
games? Percept Mot Skills. 84, 3-15.
Smith, P. J. and Davies, M., 1995. Applying contextual interference to the Pawlata roll.
J Sports Sci. 13, 455-462.
Tanji, J., 1996. New concepts of the supplementary motor area. Curr Opin Neurobiol.
6, 782-787.
Taylor, J. L., Wagener, D. S. and Colebatch, J. G., 1995. Mapping of cortical sites
where transcranial magnetic stimulation results in delay of voluntary
movement. Electroencephalogr Clin Neurophysiol. 97, 341-348.
Thach, W. T., 1997. Context-response linkage. Int Rev Neurobiol. 41, 599-611.
Thickbroom, G. W., Byrnes, M. L., Sacco, P., Ghosh, S., Morris, I. T. and Mastaglia,
F. L., 2000. The role of the supplementary motor area in externally timed
movement: the influence of predictability of movement timing. Brain Res. 874,
233-241.
Thomas, J. R., Salazar, W. and Landers, D. M., 1991. What is missing in p less
than .05? Effect size. Res Q Exerc Sport. 62, 344-348.
Townsend, B. R., Paninski, L. and Lemon, R. N., 2006. Linear encoding of muscle
activity in primary motor cortex and cerebellum. J Neurophysiol. 96, 2578-
2592.
87
Tsutsui, S., Lee, T. D. and Hodges, N. J., 1998. Contextual interference in learning
new patterns of bimanual coordination. J Mot Behav. 30, 151-157.
Van Donkelaar, P., Lee, J. H. and Drew, A. S., 2000. Transcranial magnetic
stimulation disrupts eye-hand interactions in the posterior parietal cortex. J
Neurophysiol. 84, 1677-1680.
van Mier, H., Tempel, L. W., Perlmutter, J. S., Raichle, M. E. and Petersen, S. E.,
1998. Changes in brain activity during motor learning measured with PET:
effects of hand of performance and practice. J Neurophysiol. 80, 2177-2199.
van Mier, H. I., Perlmutter, J. S. and Petersen, S. E., 2004. Functional changes in brain
activity during acquisition and practice of movement sequences. Motor Control.
8, 500-520.
Wassermann, E. M., 1998. Risk and safety of repetitive transcranial magnetic
stimulation: report and suggested guidelines from the International Workshop
on the Safety of Repetitive Transcranial Magnetic Stimulation, June 5-7, 1996.
Electroencephalogr Clin Neurophysiol. 108, 1-16.
Weekley, A., 2004. TEMPLATE). Weekley Consulting, Garden Grove.
Werhahn, K. J., Fong, J. K., Meyer, B. U., Priori, A., Rothwell, J. C., Day, B. L. and
Thompson, P. D., 1994. The effect of magnetic coil orientation on the latency
of surface EMG and single motor unit responses in the first dorsal interosseous
muscle. Electroencephalogr Clin Neurophysiol. 93, 138-146.
Willingham, D. B., 1998. A neuropsychological theory of motor skill learning.
Psychol Rev. 105, 558-584.
Wise, S. P., Moody, S. L., Blomstrom, K. J. and Mitz, A. R., 1998. Changes in motor
cortical activity during visuomotor adaptation. Exp Brain Res. 121, 285-299.
Wright, D. L., Li, Y. and Coady, W., 1997. Cognitive processes related to contextual
interference and observational learning: a replication of Blandin, Proteau and
Alain (1994). Res Q Exerc Sport. 68, 106-109.
Wright, D. L., Li, Y. and Whitacre, C., 1992. The contribution of elaborative
processing to the contextual interference effect. Res Q Exerc Sport. 63, 30-37.
Wright, D. L., Magnuson, C. E. and Black, C. B., 2005. Programming and
reprogramming sequence timing following high and low contextual
interference practice. Res Q Exerc Sport. 76, 258-266.
88
Wrisberg, C. A. and Liu, Z., 1991. The effect of contextual variety on the practice,
retention, and transfer of an applied motor skill. Res Q Exerc Sport. 62, 406-
412.
Wu, A. D., 2004. DataWizard), Los Angeles.
Wulf, G., 1992. The learning of generalized motor programs and motor schemata:
effects of KR relative frquency and contextual interference. J Human
Movement Studies. 23, 53-76.
Wulf, G. and Lee, T. D., 1993. Contextual interference in movements of the same
class: differential effects on program and parameter learning. J Mot Behav. 25,
254-263.
Wulf, G., Schmidt, R. A. and Deubel, H., 1993. Reduced feedback frequency
enhances generalized motor program learning but not parameterization
learning. J Exp Psychol Learn Mem Cogn. 19, 1134-1150.
Young, D. E., Cohen, M. J. and Husak, W. S., 1993. Contextual interference and
motor skill acquisition: on the processes that influence retention. Human
Movement Science. 12, 577-600.
Zhang, J., Riehle, A. and Requin, J., 1997. Analyzing Neuronal Processing Locus in
Stimulus-Response Association Tasks. J Math Psychol. 41, 219-236.
Ziemann, U., 2004. TMS induced plasticity in human cortex. Rev Neurosci. 15, 253-
266.
Ziemann, U., Hallett, M. and Cohen, L. G., 1998. Mechanisms of deafferentation-
induced plasticity in human motor cortex. J Neurosci. 18, 7000-7007.
Ziemann, U., Ilic, T. V., Pauli, C., Meintzschel, F. and Ruge, D., 2004. Learning
modifies subsequent induction of long-term potentiation-like and long-term
depression-like plasticity in human motor cortex. J Neurosci. 24, 1666-1672.
Abstract (if available)
Abstract
This dissertation was designed to investigate the neural basis implementing the contextual interference effect in motor skill learning. Sixty-one non-disabled adults were recruited. Participants practiced three fast, discrete, goal-directed arm movements each with specific time and amplitude requirements. The motor tasks were practiced either in a blocked or quasi-random order. Transcranial magnetic stimulation (TMS) was applied to the arm areas of the cortical motor system (CM) to directly perturb brain processing during motor practice. Single TMS pulses were delivered, synchronized to each inter-trial interval. The three stimulation conditions (no TMS Control, TMS, Sham) and two practice orders (Blocked, Random) factorial design resulted in six experimental groups. Testing took place over 2 consecutive days with acquisition and immediate retention phases on day 1 and a delayed retention phase on day 2. The retention tests consisted of trials without feedback with which neither TMS nor Sham-TMS applied.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Computational transcranial magnetic stimulation (TMS)
PDF
Neural substrates of motor memory consolidation: a double dissociation of primary motor cortex and dorsolateral prefrontal cortex induced by practice structure
PDF
The brain and behavior of motor learning: the what, how and where
PDF
Neural substrates associated with context-dependent learning
PDF
Development of automaticity for discrete movements: A systematic investigation of the dual-task probe paradigm
PDF
Relationship between interhemispheric inhibition and bimanual coordination in musicians
PDF
Learning reaching skills in non-disabled and post-stroke individuals
PDF
Corticomotor excitability of gluteus maximus: influence on hip extensor strength and hip mechanics
PDF
Hemisphere-specific deficits in the control of bimanual movements after stroke
PDF
The footprint of pain: investigating persistence of altered trunk control in recurrent low back pain
PDF
Trunk control during dynamic balance: effects of cognitive dual-task interference and a history of recurrent low back pain
Asset Metadata
Creator
Lin, Chien-Ho
(author)
Core Title
Contextual interference in motor skill learning: an investigation of the practice schedule effect using transcranial magnetic stimulation (TMS)
School
School of Dentistry
Degree
Doctor of Philosophy
Degree Program
Biokinesiology
Publication Date
07/31/2007
Defense Date
05/30/2007
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
contextual interference,motor learning,OAI-PMH Harvest,TMS
Language
English
Advisor
Winstein, Carolee J. (
committee chair
), Azen, Stanley Paul (
committee member
), Gordon, James (
committee member
), Sullivan, Kathy (
committee member
), Wu, Allan (
committee member
)
Creator Email
chienhol@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-m728
Unique identifier
UC1220942
Identifier
etd-Lin-20070731 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-534922 (legacy record id),usctheses-m728 (legacy record id)
Legacy Identifier
etd-Lin-20070731.pdf
Dmrecord
534922
Document Type
Dissertation
Rights
Lin, Chien-Ho
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Repository Name
Libraries, University of Southern California
Repository Location
Los Angeles, California
Repository Email
cisadmin@lib.usc.edu
Tags
contextual interference
motor learning
TMS