University of Southern California Dissertations and Theses

Functional water MR spectroscopy of stimulated visual cortex using single voxel

(USC Thesis Other)

Functional water MR spectroscopy of stimulated visual cortex using single voxel

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content FUNCTIONAL WATER MR SPECTROSCOPY OF STIMULATED VISUAL CORTEX USING SINGLE VOXEL, by Pankaj B. Patel A Thesis Presented to the FACULTY OF THE SCHOOL OF ENGINEERING UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree MASTER OF SCIENCE IN BIOMEDICAL ENGINEERING August 1994 Copyright 1994 Pankaj B. Patel UMI Number: EP41387 All rights reserved INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted. In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion. Di sser t at i on Pu bl ishing UMI EP41387 Published by ProQuest LLC (2014). Copyright in the Dissertation held by the Author. Microform Edition © ProQuest LLC. All rights reserved. This work is protected against unauthorized copying under Title 17, United States Code ProQuest LLC. 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, Ml 48106- 1346 T h is thesis, w ritte n by .Pankaj B .t P. a. t e. l..................................................... under the guidance of h is F a cu lty Comm ittee and approved by a ll its members, has beeti presented to and accepted by the School of E ngineering in p a rtia l fu lfillm e n t o f the re quirements fo r the degree of Master of Science Date...... 07 / 28 / 94 Faculty Committee _ Chairman Dedication This Thesis is dedicated for the pleasure of Lord Krishna , for my parents and lastly for my beloved wife. n Acknowledgements: I would like to thank all of those who helped and guided me to complete this project for my thesis. First of all, I would like to thank my advisor and my committee chairman, Dr. Manbir Singh, who provided me a lot of guidance about the experiments and analysis, without which I would have not been able to complete this project. I would also like to extend my gratitude to Mr. Deepak Khosla, who helped me in processing and programming for data analysis. I am also grateful to Mr. T. S. Kim and Mr. Hyun Kim, whose help was appreciable in setting up the experiments at MRI Imaging Center of USC County Hospital. m Table o f Contents Topic: Page: 1 - Introduction 1 2 - Review of Magnetic Resonance 4 -Principle of MRI 5 -Acquiring the Signal 8 -FID 10 -Gradient Coils and VOI -Slice Selection and ROI 11 -MR Parameters 12 .T1 Relaxation 14 .T2 Relaxation 16 -Pulse Sequences 18 -Magnetic Resonance Spectroscopy 20 3 - Functional Water Spectroscopy of Stimulated Visual Cortex 28 -Methodology 29 -Results and Discussion 32 -Conclusion 44 -Appendices : Appendix : A 46 Appendix : B 53 4 - References 55 IV List o f Figures Title Page fig.l Behavior of atomic nucleus 5 fig.2 Parallel and anti-parallel spin 5 fig.3 Precession of nucleus 7 fig. 4 Alignment of proton 7 fig.5 Energy states under magnetic field 8 fig. 6 Transverse Magnetization 9 fig. 7 Effect of transverse magnetization 9 fig.8 Signal capturing 9 fig. 9 FID and its fourier transform 10 fig.10 Effect of field gradient 11 fig.11 Slice Selection 13 fig.12 Spin-lattice relaxation 15 fig.13 Magnetization Recovery 15 fig.14 Effect of T1 and T2 parameter 16 fig.15 Spin-Echo 17 fig.16 Field inhomogeneity and T2* 19 fig.17 Water and fat in spectroscopy 23 fig.18 Voxel localization 31 fig.19 Functional MR image 31 V In tro d u ctio n : The study of medical imaging is related to interactions between tissue and/or bone and all forms of energy. Observing these interactions has evolved and major progress and breakthrough have been made since 1895. The first imaging technique was concerned with the invention of X-ray by a scientist Rontegen . The first image was reported by the German scientist "Guerd", when he tried to image the forearm of his wife using a primitive X- ray technique developed by him at that time. He found out that he could take images of the bones, but not that of the soft tissues in the human body. Using the regular X-ray technique, one could get two dimensional images, which proved to be a limitation to giving proper patient diagnosis. Therefore, the need to get better imaging technique was necessary, and scientists worked hard throughout the twentieth century and especially in the last three decades to invent a technique that gave three dimensional X-ray images. The latter was first applied in England in 1972, and since then the technology has evolved Computer aided Tomography (CAT). Biological tissue is relatively transparent to X-rays and opaque to radiation with intermediate wavelength. There is a window in tissue absorption through which radio waves can be used to probe deep inside the human body.Many other 1 scientists have been working in developing a non invasive technique to image both soft and hard tissues in the body. This technique is known as Nuclear Magnetic Resonance (NMR) where benefits derived from the use of low energy radiation and the unprecedented level of information available from nuclear signals are combined to make imaging. (19) . One of the first NMR signals were produced by Jasper Jackson from a live animal in 1967 and NMR images by Lauterbur in 1973 using a modified, conventional NMR spectrometer. (4,5,6) The potential of this technique was immediately apparent, since it allowed radiologists to diagnose diseases with greater sensitivity and specificity--the NMR parameter Tl and T2 appeared to be altered in tumors when compared with corresponding normal tissues. Because of the fact that this method presents no known intrinsic hazard to prospective patients, and since these new images were in some ten order of magnitude in terms of quality and resolution better than the known X-ray images, advances in the methods of applications and instrumentations were introduced to allow the technique to be used in clinical trial worldwide.(4,5,6) In this thesis, I am mainly concerned with Magnetic Resonance Imaging(MRI) , specifically, Magnetic Resonance Spectroscopy (MRS). I start by a review of the principles of the magnetic resonance imaging -- the physics related to 2 it, the parameters Tl and T2, signal sequences etc. Then I will discuss about Magnetic Resonance Spectroscopy . The objective of this research is to complement local change in blood flow in concerned cortex (visual cortex) and so change in the MR parameter T2 as well as a very small change in proton abundance under stimulus of flashing light on checker board in front of eyes.(1,2,3) In this Part, I will present some literature review, and then I will go into the methodology of my research, show the results and discuss them, and finally, conclude with my remarks about the future of the field of MRS. 3 Review o f Magnetic Resonance Im aging : NMR (Nuclear Magnetic Resonance - as also called) images were first taken from a live animal before they tested them on humans, and the first one who obtained NMR signals from humans was Damadian in 1971.(4,6) Then, two- dimensional images of a water sample were taken by Lauterbur in 1973, and since then significant developments have been made in the application of NMR spectroscopy in the clinical setting for the study of biological systems. (12) We all know that soft tissues are transparent to the regular X-ray technique, and that they are not differentiated nicely in X-ray CT due to the high energy state of the x-ray technique. MRI allowed us to access information about biological tissues within intact biological samples which is almost unavailable to us with the usage of other techniques. The advantages of MRI are that it does not have any ionization radiation, has excellent tissue differentiation (due to different MR parameters of different tissues), has an inherently three- dimensional capability, has very easy and fast reconstruction technique, has a very flexible system of imaging, and give us information helpful not only for diagnosis but also for the study of biochemical reactions in the human body.(13,16,18,20,28) 4 Principles o f M R I: Nuclei are the tiny cores of the atoms which make up the elements of our universe. Certain nuclear species possess angular momentum or spin, first suggested by Austrian physicist Wolfgang Pauli. He mentioned that charge bearing nuclei behave like a minute magnet and therefore produce a magnetic moment mu, as shown in the fig.1.(19) Parallel Antiparal (a) fig.2 When immersed in a static magnetic field, the randomly oriented magnetic dipoles (fig. 2a) respond to the force of 5 the field and try to get aligned with it.(fig 2b) Pauli discovered that certain nuclei with an odd number of protons possess a relatively high magnetic moment, and it is this particular property that is utilized in imaging these nuclei. Such nuclei are the proton 2H(the principal isotope of hydrogen), ^C, ^Li, 18p^ etc., but in this paper, I will be looking only at the proton because of its abundance in the human body. The protons keep on spinning and the spins are randomly oriented in the absence of an external magnetic field, Bo. But, with the presence of a BO, there are two allowable states with orientations "parallel" or spin up and "anti-parallel" or spin down corresponding to high and low energy states E2 and El as shown in the fig.2 above.(19,20) We see from the figure above that nuclei are actually alined at an angle theta to Bo. Therefore, the nuclei has a spinning motion around its axis and a precession motion around the Bo magnetic field.(fig.3,4) The presence of Bo produces a gravitational torque: — suxBo dt then, — =/I x rBo dt where, T is the Gyromagnetic ratio, usually a constant in (MHz/Tesla). (19,20,24,25) 6 Precession Motion Spinning Motion Gravitation Br A _ _ _ fig-3 fig, 4 So, we end up having a relationship between the angular frequency (Do , and the magnetic field Bo called the Larmor Frequency and is given by the following: COO = pBO The proton (1H), the principle isotope of hydrogen gives us the best signals for many reasons (19,20): (1) its isotopic abundance is about 99.98% (2) it has the highest chemical abundance in the body in the form of water and lipids. (3) it has unity relative sensitivity. (4) it has the highest magnetic moment amongst all other nuclei (Magnetogyric Ratio ~ 42.58 MHz/Tesla) The individual spin vectors— both parallel and anti parallel— make up a net magnetic moment M along the z-axis, Since we have two different stages, one is parallel (low 7 energy) and the other one is anti-parallel (high energy), resonance is the difference of energy between the two stages as shown in the figure below:(fig. 5) As I discussed, the net magnetic moment M due to nuclei is also in the Z-direction, the direction of the main huge magnetic field of the system, Bo. Hence it is not as simple to measure this dominated signal from nuclei. To acquire this very weak signal in presence of strong magnetic field, it has to be rotated in transverse plane somehow. The transverse signal will be precessing in the x- y plan, thereby inducing an A.C. signal that could be captured with a receiver coil. (7,19,20,21,22) E_» (Spin down, high energy) E+3 " 2 (Spin up, low energy) fig.5 Acquiring the Signals : — w + m m m m i^— —— —— 8 The process of transversing is done by charging the protons with energy of amplitude Bl in the x-y plan, the latter should be rotating at an exact frequency as the proton precession frequency, (fig. 6,7,8)(19) 9 F ID : Now RF pulse is applied only to transverse the signal.Hence Mxy dyes out after the RF pulse is turned off, therefore the voltage received through the coil decays to zero and has the form of a damped oscillation shown below. The signal is herein known as Free Induction Decay (FID) representing the time evolution of the transverse magnetization. We then have to transform this time domain signal into a frequency domain signal, by taking its Fourier Transform (F.T.). (fig.9) The F.T. allows us to extract the individual frequencies as well as their corresponding amplitudes (which is proportional to corresponding contents of that nucleus). (12,13,14,19,20,21,26,27) (a) Time Frequency /VV-MAjVv-w fig. 9 10 Gradient Coil and Volume o f Interest As shown in fig.10, it is quite impossible to localize the content of nuclei in volume of interest without giving any gradient of magnetic field along X and Y direction. Due to different magnetic field, there will be different resonance frequency(which is proportional to magnetic field) and so in frequency domain different frequency will indicate the location. If we make the magnetic field to vary linearly with a determined axis, the exact location of the volume of interest or VOI would then be easily located. *vVWv'"'W Time Time Slice Selection and Region o f Interest (ROI) : In order to get a cross - sectional image (as Most MRI images are shown) one has to constrain excitation to a thin slice. There is a technique with which we can confine the data acquisition from thin slices of region of interest or ROI, this technique is called Selective Excitation of a Slice. To select a slice, we have to make the radio frequency excitation itself spatially-selective. We need to design and use a spatially selective RF pulse--usually the range of the pulse range from 1 to 2 Larmor frequencies. For subjects lying down along the z-axis of the magnetic field, we also have to employ with the pulse, a z- gradient Gz, that limits the excitation of the protons to regions of interests in a slice z. The band of frequency of excitation pulse, along with the gradient confines excitation of slice, no signals will be excited or detected outside the defined slice.(fig.11)(19) To find the RF pulse that will excite a narrow band of NMR frequencies in the range CO 1 to CO2, we use the function rect ( co ) =1, I co I <1/2; = 0, I fill > 1/2; 12 The Fourier Transform FT{rect(&?)} gives a sine RF pulse function that is used for selective slice excitation along with Gz gradient. fig.11 13 The M R Parameters: The NMR signals picked up by the receiver represent those signals coming form the soft tissues protons. The protons of protein, DNA, and solid structures like the bone do not give out signals, therefore they appear black in MRI. (Bones can't be imaged by MRI - which may be considered as the greatest weakpoint of MRI) The main advantage of MRI over other imaging techniques is the relative contrast of the images, and that is due to two determinants Tl (spin-lattice or longitudinal) relaxation time, and T2 (spin-spin or transverse) relaxation time. (4,5,6,14,18,19,20,21,22,28) T l Relaxation: After the RF pulse, the protons are in an excited or high energy state, the nuclei are considered hot, so they dissipate their excess energy to the lattice that is considered to be cold, and that is why it is known as the spin-lattice relaxation time. The energy dissipation time is equivalent to the time for the transverse magnetization to get restored to its original orientation along the z- axis. The restored magnetization has a recovery rate of 63% after 1*T1) time, 86% after 2*T1 time, and 95% after 3*T1 time, hence in practice we usually use 4*T1 time to recover the entire magnetization(fig. 12,13,14) (18,19,20). 14 (18,19,20) It is also known that Tl is dependent on molecular size and their corresponding tumbling rate, for the case of pure water (small molecules) the Tl is about 3 seconds, while for fat (larger molecules) the Tl is in the range of few hundred milliseconds, and so, Tl is quite different for large (short Tl) and small molecules, (long Tl) (4,6,18,19,20) Spins ( h o t ) Lattice (cold) fig.12 - M T, 3T, fig.13 In short, the Tl is the measure of how long it takes the transverse magnetization to achieve its initial value M along the z-axis. 15 T2 Relaxation: This relaxation is also known as spin-spin relaxation time, because in the process energy is not only transferred from the nuclei to the lattice, but also energy is transferred amongst nuclei. I mentioned earlier that the transverse magnetization decays to zero because the magnetic moment get out of phase, and T2 measures that rate of decay, (fig.14). X* X* fig.14 16 Different nuclei precess at different frequencies due to the fact of the inhomogeneity of the external magnetic field, which lead to the loss of phase coherence with losb of transverse magnetization. In the case of field homogeneity, the decay would be equal to T2, and since we clearly have inhomogeneity, the decay time constant would be T2*, which is shorter than the ideal T2.(fig.l5) Therefore, data acquisition lasts for T2 period of time. T2 is more efficient than Tl, and hence, is less susceptible to the magnitude of the external field (long T2 for small molecules and short T2 for large molecules) . (6,18,19,20,21,22) Tl and T2 occur in parallel and contribute simultaneously to the NMR signal detected by the receiver coil which can be used for any kind of MRI in specific ways.(7,8,9,10,11) Detection decay ~ T 2 decay Spin Echo r— 180 fig.15 17 Pulse Sequences : There are different pulse sequences being used for different type of applications.The most useful pulse sequence is spin-echo technique. (5,18,19,20) -phe initial transverse magnetization decay, and loss of coherence or dephasing caused by field inhomogeneity could be reestablished by applying a 180 pulse at t = Tj seconds after the initial 90 pulse. This method turns the partially dephased magnetization into a mirror image position, thus leading to a refocusing at an echo time 2Tj, also known as the echo delay Tg. (fig.14) Both, Tl and T2 contribute to the signal intensity S given by the following relationship: S = exp(-£)x{l-exp(-£)} Where, Tr is the repetition time. And, the spin-echo sequence can be summarized as: (90 - Tp - 180)n Here, n could be larger than one, which means that we can apply more than one echo (in practice, we usually use up to four echoes), leading to more than one image obtained from the same anatomical region each with different T2 weighting which are averaged to create final image. (18,19,20,21). 18 The fig.16 show the time sequence for the spin-echo Magnetic Resonance Spectroscopy (MRS): The physics of MRS is the same as those of MRI, But until recently MRS was very limited in its capabilities due to the limitations of the spectrometers. In the last decade, lots of technological advancements in electronics, microcomputers, superconducting magnets that maintain a high field homogeneity have made it easier to control for MR sensitivity by the order of many magnitudes. (4,6,15,16,17,18,20) MRS is also based on the fact that nuclei with an odd number of protons have a magnetic moment due to its precession around its axis, and that these nuclei align either in parallel or anti-parallel with the external magnetic field. Also, MRS is based on the fact that surrounding electron clouds of adjacent molecules tend to shield the penetration of the external magnetic field, and therefore, produce a chemical shift that depends on the size and site of the molecules. (15,16,18,20) The external field (Bo) induces an electron current that generates a local field (Bj_oc) in opposite direction to the original field. The sum of these two field gives the nuclei chemical shift resonance frequencies, which are different from one nuclei to another. This difference in frequency is essential in obtaining information in MRS, because it is through these differences the one can study 20 and characterize the chemical constituents of the tissues being imaged. (18,20) Spectroscopic information is usually generated in form of spectra consisting of plots of signal intensity versus resonance frequencies in a homogeneous field using a standard chemical compound as the reference. Shifts in resonance frequencies are expressed in parts per million or ppm, which makes the chemical shifts independent of the field magnetic strength. Studying the spectra carefully can give us important information about the concentration of the nuclei, we get that simply by calculation of the areas under the spectral peaks.(18,20,29) In clinical applications, we usually get these signals from localized regions either in two or three dimensions. Human tissues consist mainly of water and fat that are separated by 3.5 ppm, and so, the simplest format of protons imaging is the separation of water and fat. The process of chemical shift is expressed by the following : Bloc=B° + Bint where, Bo, is the external magnetic field Bloc' the local magnetic field felt by the nucleus Bint, is the internal magnetic field generated near the nucleus and is equal to: - ( <T*Bo) , 21 where O' is a constant, so then Bi0c — ^ ® ) Bo The resonance frequencies for two different nuclei of the same species are: uref = (V/2x) * (1—crref ) Bo where, ref stands for reference t>S = (U/2JT) * (1-O-g) Bo where, s stands for sample The definition of the relative chemical shift in parts per millions is given by : S = 10 ^ (l)s_ t)ref ) / vref MRS main purpose is to provide metabolic information due to biological and chemical system changes due diseases and tumors. Due to the high abundance and high concentration of water and lipids in human tissues, the intensity of the signals in their corresponding spectra give high quality images with excellent spatial resolution. The separation of fat and water could lead to important implication about certain diseases and tumors, and the separation of less abundant compounds and metabolites— in the range of 4 order of magnitude lower than that of tissue water— like N-acetyl-aspartate (NAA) and Lactate (Lac) give greater metabolic information. Proton MRS is one of the 22 first methods to be used in clinical areas.Fig. 17 a,b shows MRS spectra for different chemicals. Fat (Q H 12) Water (H2O) PCr P-ATP Y-ATP a-A T P PDE (b) PME fig.17 Water Imaging: The process is done by three methods, l)spin-echo sequence; 2) gradient echo sequence; and 3)frequency selective pulse.(18,20) l)The spin-echo sequence uses a 180 RF pulse shifted from the middle of the sequence. This methods depends on the fact that water and fat signals either add or subtract depending on the phase of the signals. In regular spin-echo sequences the two signals add, but in the case we shift the RF pulse by a time t, 23 t=l/ (2 Af) where, A = difference in resonance frequency between fat and water. In this case we have the fat and water signal 180-degree apart, and thus their corresponding signals cancels. Water and fat images are produced after the combination of phase and magnitude data. 2)The gradient-echo sequence uses a gradient-echo pulse with specific echo time to cause the signal to either add or subtract. Water and fat images are obtained by combination on gradient echo images. 3)The frequency selective pulses employs specific pulse that either excite or saturate the nuclei according to their chemical shift selective [CHESS] saturation. Signals from water and fat are eliminated when imaging sequence is preceded by a single frequency saturation pulse. Localized Spectroscopy: The first step to spectroscopy experiment is to define the volume of interest. (15,18) The importance of successfully localizing the volume of interest (VOI) cannot be stressed enough, because imperfect localization could lead to contamination from the surrounding fat tissues, thus leading to faulty data. There are many methods to do the perfect localization and many other methods are under 24 development. Some of these methods are classified by either using Bo method--that rely on the application of magnetic field gradient for spatial localization-- or by using B1 method--that rely on using transmitter coil for spatial localization or a combination of both.(18) The Bo approaches include Topical Magnetic Resonance (TMR), Stimulated-Echo Acquisition Method (STEAM); Image Selected In Vivo Spectroscopy (ISIS), and Spatially Resolved complicated echo sequences. There are also some methods that employ a combination of both BO and Bl approaches like Depth Resolved Surface Coil Spectroscopy (DRESS) and Fast Rotating Gradient Spectroscopy (FROGS). Surface Coil Localization: This method confines the data collection to areas close to the coil. The RF fields arising from the surface coil is highly inhomogeneous, and so we can manipulate the region from which signals arises. The figure below shows how the signals intensity very with respect to depth, and therefore with respect to closeness of the VOI to the surface coil. Intensity is at the highest when the depth is minimal. 25 In the study of this thesis, the standard head coil of a 1.5 T Philips Gyroscan was used for exciting and receiving the signal. These localization techniques are very important to study the specific chemicals. Localization techniques take good part of the study time, which makes impractical for rapid clinical through-put in a busy hospital or MRI center. The optimal way to study human metabolism is done nowadays in research setting because the study require repeated patient studies and consumes expensive machine time.This is called setup time which may vary according to the expertness of user, but improvement in the localization techniques can definitely decrease this dead time and we can get to a point where the process takes less time, thus widening its usage in clinical setting and finally, leading to more research in the area of MRS of water and metabolites. (29,23) Magnetic Field Inhomogeneity and Shimming: As we discussed earlier, field inhomogeneity affects the transverse magnetization decay (fig.17), and the more inhomogeneous the field the faster the decay time T2*, which broadens the spectra and reduces the peak height. We usually limit the voxel size to the range of 1-10 cm, and it is required to have a uniform static magnetic field over 26 the specified voxel in order to obtain high resolution spectra. (23) Since biological objects are inhomogeneous, we can correct for the problem by shimming the magnet so that the inhomogeneity are compensated over the specified voxel. (29, 23) The shimming process is done by adjusting the magnetic field shim settings along the orthogonal axis and along combinations of the axis. The local magnetic field is adjusted by changing the current amplitudes to a shim coil set within the magnet bore. The shim coil are designed to correct for linear errors along the three orthogonal axis.(23) These shimming adjustments can be easily changed with the study design, and they are usually incorporated in the MRI control console. 27 Functional Water Spectroscopy o f Stimulated Visual Cortex The NMR spectroscopy or MRS of the brain during ON and OFF the stimulus is relatively a new and rapidly growing field of research. Because of the non-invasiveness of the MRS method, presenting no risk of harm to humans, MRS allows measurements to be performed repeatedly on the same subject as often as necessary, yielding important and valuable metabolic information unavailable to us otherwise. (18,20) This is why MRS is overcoming other imaging modalities. In this thesis, I am mainly interested in a very simple and most common component of human body and that's WATER. In brain, almost all the compartments like blood vessels, cerebrospinal fluid (CSF), brain tissue etc. contain water. Hence simply by water spectroscopy, many unknown and surprising information about brain activities under certain kind of stimulation can be predicted. As I mentioned, spectroscopy analysis gives information about the content of certain chemical (here water) as well as change in inhomogeneity. It is well established that during stimulation, blood flow in related areas increases and at the same time oxygen consumption does not increase. This kind of change leads to decrease in deoxyhemoglobin which results in increased 28 magnetic field homogeneity in local areas. This proves an increase in spin-spin time constant T2*.(l,2) Here T2* ( also called decay constant ) increases which results in an increase of peak height. Now there may be an increase in area of the peak. If it is, then it also may cause an increase in peak height.(3) With this expectation, we started our experiments. Methodology : We did experiments on seven subjects (30) including myself. We also repeated same experiments on same subjects to support strongly our results. We had imaged under a 1.5T Philips Gyroscan MRI/MRS system at health science campus, USC. In our setup, we arranged a big Checker board in front of the eyes of the subjects. To provide a stimulus to visual cortex, we used a flash light exposing on checker board. The flashing light was being controlled from control room. To avoid the effect of external sun light, we always performed experiments after it gets dark enough and still the subject was asked to keep his eyes closed during OFF stimulus data acquisition. MRI and MRS were combined in the following imaging process: 1) scout images of the entire head were first taken and the process took about 2 minutes. 29 2) a single voxel of (30x15x30 cc) was used in this study. Sagital and coronal scout images positioned the voxel to enclose the visual cortices. This was done with great care to prevent contamination from surrounding regions. Fig.18 shows the area of the brain where we located our single voxel. 3) head coil was selected to help in localization, and parameters were changed to fit the study. VOI and echo were both enabled, 8 averages were chosen, and 50 ms was the echo time. This echo time was seemed to be optimum. The repetition time was 4000 ms. 4) shimming was then performed to compensate for the magnetic field and compounds inhomogeneity. 5) gradient adjustment was then performed. 6) step 4 and 5 were repeated twice or sometimes thrice unless we had met with good shimming. 7) data ( see appendix B) were acquired for a single voxel with 5 to 7 sets. Each set includes one ON and one OFF data. 8) we did collect data for controlled case in all subjects.(which means data from unrelated region to see the obvious effect of stimulus on related areas of the brain.) 9) we also did functional MRI to make sure our localization of single voxel. Fig.19 shows the cross 30 sectional image of functional MRI. The activated area was our site for single voxel. fig.18 (a) fig.18 (b) fig.19 10) all collected data were stored and then processed at main campus, USC. The software package mainly used for processing was MATLAB (version 4.1). 11) here our single voxel was on inclined transaxial plane and so the processing was not as simple as it is with horizontal transaxial plane. We met with certain artefact 31 from gradient coil. I spent considerable time to get rid of that by processing. Finally I could analyze data in both time as well as frequency domain and thus I had cross-check on results. In time domain, I used exponential curve and in frequency domain lorentzian curve for curve fitting minimizing least square errors. We also made sure with Prony's technique for analysis. Results And Discussion: First of all, I would like to present some results from i r r y analysis method. Fig. r.l and r.2 shows the plots of original data and best fit (using least square approach) in time and frequency domain respectively. In r.3 , individual components of best possible fit in frequency domain are shown. N.B.:-fig.r.? indicates result graph. Solid plot indicates fitted curve and dashed plot indicates original data. Here the results from two subjects are presented. To observe the effect of stimulus on T2*, peak height of spectra and area of the spectra, they are normalized to proper scale. Fig. r.4(a),(b) represents the corresponding change in all three parameters in active and controlled region 32 respectively. While in fig.r.5,6,7 active and controlled data set for each parameter are shown individually. Here as it is obvious that there is a consistent change in T2* and peak height of the spectra and very small change in area of the curve. Fig.r.8,9 presents the same parameters during ON and OFF stimulus for other subjects.Other plots for remaining data sets are presented in appendix A. All the data sets and percentage changes are shown in the table I and overall changes are shown in table II in appendix B. We saw three obvious results. One is increase in T2* (by an average of 1.710 %) that causes a decrease in full width half max (FWHM) in fourier domain. The second one is is an increase in peak height (by 2.171 %) and the third is a little increase in area (by 0.2143 %). The decrease in fwhm results in peak height increase. Also the increase in area causes increase in height. The area of the peak in fourier domain represents the number of total proton (which is almost proportional to water content) in single voxel. Thus during stimulation, spin-spin relaxation time increases abundance of proton may also increase slightly. 33 F ID (modulus) dotted: original data, solid: best f i t 600 500 400 300 200 100 1200 1000 800 600 time in mSec 200 400 fig.r.l 34 F T o f F ID (modulus) dotted: original data, solid: best f i t 2.5 1.5 0.5 100 150 200 -100 -50 freq. in Hz f i g . r . 2 35 F T o f F ID (modulus) dotted: original data, s o l i d : individual compo. of best f i t 2.5 0.5 200 150 100 50 50 -100 -50 freq. in Hz fig.r.3 36 H t on Data from "Sub 1" 60 Area T2 50 - 40 - 30 - Ht. -a- 20 0 5 6 2 3 1 4 Ht on Ht off T2 on T2 off Aron Aroff # of set fig.r.4.(a) 37 H t on Data from "Sub 1 cont" 70 12 60 - Area 50 - 40 - 30 - Ht. -o- 9= 20 0 6 8 2 4 Ht on Ht off T2 on T2 off Ar on Ar off # of set fig.r.4.(b) 38 H t on Data from "Sub 1" 24 23 - 22 0 5 6 1 2 3 4 Ht on Ht off f Of M t Data from "Sub 1 cont* 26 25 24 23 0 8 4 6 2 -a Ht on '• ■ ' Htoff # of *«t fig.r.5 39 T 2 o n T 2 on Data from "Sub 1" 50 48 - 47 0 2 3 4 5 6 1 # of sot Data from "Sub 1 cont" 63 62 - 61 - 60 6 8 0 2 4 # of «ot f i g . r .6 40 A r on Data from "Sub 1" 58.4 58.2- 58.0- 57.6 0 1 2 3 4 5 6 * of s«t Data from "Sub 1 cont" 53.0 52.8- 52.6- 52.4- 52.0 0 6 2 8 4 # Of M t fig.r.7 41 H t on Data from "Sub 2" ■ a Ht on * ----- Ht off a T2 on ■ e T2 off ■ ■ ----- Ar on a Ar off 0 1 2 3 4 5 6 7 # of set fig.r.8 Area 42 Data from "Sub 2" e a o t 5 2 3 4 0 7 T2on T2ofl • Of M t Data from “Sub 2“ sa.2 88.0 57.8 7 0 s 8 1 2 3 4 Aron Aron Data from "Sub 2" Won Hon fig.r.9 43 Conclusion: I have successfully integrated the MRI/MRS concepts to generate new information. Still it's a long way to go so that we can achieve even better result. There will definitely be some hardware development, giving us better RF coil, better MR spectrometer, and may be more powerful magnets and thus higher magnetic field. Higher magnetic field improves the MRS process by improving the water- suppressed proton spectroscopy. There are some centers that employ an open MRI Imaging system that has minimal gradient sound. These systems may be the trend for the future and for the development of even more advanced techniques. Some improvements could be done in the localization methods. If we improve on the voxel selection process, it may be possible to define a smaller voxel depicting more localized spectra and information. The MRS modality is still being improved every day, and it is far from being perfect. The experiments that were performed took about one hours to finish, and due to the expensive machine time this technique will remain under the flagship of research until some revolutionary development in the field are obtained to cut the setup time. 44 In summary, I was successful in achieving my goal, by detecting a small but noticeable changes in effective spin- spin time constant (T2*) and peak height of spectra in visual cortex during visual stimulation. 45 H t on A p p e n d ix : A Data from "Sub 3" 80 T2 70 - 60 - Area 50 40 Ht. 30 - 20 0 5 6 4 1 2 3 Ht on Ht off T2 on T2 off Aron Ar off # of set 46 Data from "Sub 3" 71.2 70.6 70.6 0 3 1 2 4 S 6 T2on 12 o« f o( ««( Data from "Sub 3" Aron AroH Data from "Sub 3" 31.0 30.4 30.2 30.0 0 6 6 1 2 3 4 HI on HI o n 47 Data from “Sub 4" M .2 0 2 3 4 S 1 6 7 Data from “Sub 4" 58 0 2 3 a 7 1 4 a MM Data from aSub 4* 20 2 * 7 0 2 1 3 a a 4 48 H t on Data from "Sub 5" 70 T 2 60 - Area 50 - 40 30 Ht. =9 20 0 6 7 1 2 3 4 5 Ht on Htoff T2 on T2 off Aron Aroff # of set 49 Data from "Sub 5* T2 on T 2 o fl Data from "Sub 5" 50.7 50.5 50.3 7 0 3 5 2 4 1 • • ( Mt Data from "Sub S" 25 27 25 7 0 2 3 S 1 4 50 Data from “Sub 6 “ so sa 57 se 5 5 0 1 2 3 5 4 e * Of M t Data from “Sub 6" 42 0 2 3 4 5 e i « of Ml Data from “Sub 6* 20 I S 18 17 0 0 1 2 3 3 4 51 H t on Data from "Sub 6" 60 Area 50 - T2 40 - 30 - Ht. 20 - 4 5 6 0 1 3 Ht on Ht off T2 on T2 off Aron Aroff # of sat 52 Snb 1 APPENDIX ht on : B ht o ff * chg TABLE t2 on 33070 31872 3.758 67.650 33144 32234 2.823 67.783 32622 32231 1.213 67.505 32948 32384 1.741 67.871 33056 31831 3.848 67.874 33028 31476 4.930 68.036 Sub2 22302 21521 3.629 47.951 21903 22125 1.003 48.077 22628 22097 2.403 48.665 23110 22042 4.845 49.184 23180 21978 5.469 48.993 Sub 3 17682 17437 1.405 38.385 16826 16999 -1.017 37.305 19023 18940 0.438 42.453 19270 18501 4.156 41.938 18931 18895 6.190 41.418 Sub 4 28237 27049 4.392 59.215 28428 27288 4.177 58.851 28024 27715 1.114 58.830 27678 27696 -0.065 58.674 27960 27403 2.032 58.332 27912 27027 3.274 58.363 Sub 5 27578 27357 0.807 66.116 27237 26250 3.760 65.640 27338 26119 4.667 65.760 27431 26438 3.755 65.906 27880 26467 5.338 66.264 27107 26682 1.592 66.188 Sub 6 30512 29751 2.557 71.141 30858 30755 0.334 71.537 30470 30651 -0.590 71.494 30591 30277 1.037 71.596 30556 30209 1.148 71.021 Sub 2 23773 24705 -3.772 60.842 (C o n tr o lle d ) 25058 24675 1.552 61.265 25767 24644 4.556 62.204 24832 24655 0.717 61.650 24152 24710 -0.225 61.150 24298 24894 -2.394 62.522 24682 24059 2.589 61.652 I t2 o ff % chg ar on a r o ff % chg 65.767 2.863 584.832 580.135 0.809 66.190 2.406 585.060 583.776 0.219 66.949 0.830 579.293 577.480 0.313 66.800 1.603 582.590 581.429 0.199 65.761 3.213 583.748 580.590 0.543 65.352 4.106 581.829 577.767 0.703 47.312 1.350 582.912 580.060 0.491 47.667 0.860 583.482 579.140 0.749 48.144 1.082 580.336 576.910 0.593 48.495 1.420 579.478 578.353 0.194 48.217 1.609 579.583 577.340 0.388 37.870 1.359 582.572 585.390 -0.481 37.925 1.634 575.324 572.634 0.469 42.109 0.816 560.156 559.894 0.047 40.817 2.746 561.993 564.958 -0.524 41.168 0.607 574.925 574.382 0.095 56.850 4.16 571.805 568.188 0.636 57.522 2.31 572.408 567.967 0.781 57.480 2.348 571.999 568.052 0.694 57.820 1.476 562.998 568.758 -1.012 57.721 1.058 571.851 567.305 0.801 57.555 1.403 570.679 567.675 0.529 64.709 2.174 503.060 506.683 -0.715 63.370 3.582 506.533 506.957 -0.836 63.780 3.109 505.680 506.781 -0.217 63.574 3.668 506.357 505.808 0.108 64.235 3.158 505.365 503.713 0.327 64.251 3.014 505.338 506.388 -0.207 71.161 -0.281 506.820 503.156 0.728 70.666 1.232 515.852 514.509 0.26 70.536 1.358 513.858 513.880 0.004 70.997 0.843 515.329 514.167 0.225 71.597 -0.804 510.135 509.477 0.129 61.506 -1.079 520.447 526.516 -1.152 61.915 -1.049 522.956 521.941 0.194 61.661 0.880 526.190 521.586 0.882 61.208 0.722 523.194 522.287 0.173 62.329 -1.891 521.932 524.790 -0.544 61.966 0.897 522.836 525.137 -0.438 60.382 2.103 529.560 527.905 0.313 53 TABLE: II Sub Ht Mean Change T2* Mean Change Ar Mean Change Sub 1 3.052 2.5035 0.464 Sub 2 2.602 1.2642 0.483 Sub 3 1.1332 0.7788 -0.0789 Sub 4 2.4873 2.1258 0.4048 Sub 5 3.3198 3.1175 -0.2566 Sub 6 0.4318 0.4696 0.2694 Global Mean 2.17 1 1.710 0 .2 1 4 3 Change 54 References 1) Hennig, J., Ernst, Th., Speck, O., et al., MRM 31:85- 90, 1994. 2) Belliveau, JW., Kennedy, DN., McKinstry, RC., et al., Science 254:716- 719, 1991. 3) Singh, M., Kim, H., Khosla, D., Kim, T., in "proceeding Soc. Mag. Reso. 1994." 4) Tropp., Characterization of MR spectroscopic imaging of the human head and limb at 2.0T, Radiology 169:207- 212, 1988. 5) Pyket & Rosen, Nuclear magnetic resonance: In vivo proton chemical shift imaging. Radiology 149:197-201, 1983. 6) Matthaei et al., Multiple chemical shift selective (CHESS) MR imaging using stimulated echoes, Radiology 160:791-794, 1986. 7) Bottomly et al. , Anatomy and metabolism of the normal human brain studied by magnetic resonance at 1.5T, Radiology 150:441-446, 1984. 8) Kwong K. et al. Dynamic magnetic resonance imaging of human brain activity during primary sensory stimulation.Proc. Natl.Acad.Sci.(USA)89,5675 (1992). 9) Turner R. et al., Functional mapping of the human visual cortex at 4 tesla using deoxygenation contrast EPI, 11th Annual meeting , SVRM,Berlin, 1992,p.304. 10) Thulborn, K., et al., Oxygenation dependance of the transverse relaxation time of water protons in whole blood at high field,Biochem. Biophys, Acts 714, 265, 1982 11) Herrmann, m., et al., Functional magnetic resonance spectroscopy, A new tool for the observation of brain activation, 11th Annual Meeting,SMRM, New York :12, 1993 12) Radda, Is it clinically practical ?, Diagnostic imaging, 61-65,Sept.1985. 13) Shaw, D., Fourier transform NMR spectroscopy, New York, : Elsevier Scientific, 1971 55 14) Farrar. T. , Becker, E., Pulse and fourier transform NMR- Introduction to theory and methods, San Diego: academic press, 1971. 15) Shaw, D. , Fourier transform NMR spectroscopy. Amsterdam: Elsevier Scientific, 1976. 16) Mansfield, P., et al., NMR imaging in biomedicine, San Diego academic press, 1982. 17) Gunther H., NMR spectroscopy-an introduction, New York, Wiley, 1980.] 18) Singh., M., Cho., Z., Jones, J., Foundation of medical imaging. (Text book). 19) NMR - A perspective on imaging.(GE Electric) 20) BME-425 (USC) Class notes. 21) Herikens et al., Spatially localized MR spectra, Diagnostic imaging, 60-65, Sept. 1986. 22) Physics and instrumentation, 270-312. 23) Wherli et al., Parameters determining the appearance of NMR images. 24) Slichter, C., Principles of magnetic resonance, vol.l, Springer series in solid state sciences, Berlin, 1978. 25) Abregam. A., The principle of nuclear magnetism. Oxford : Oxford university press, 1961. 26) Bottomly, P., NMR imaging techniques and application: a review , Rev. Sci. Instr. 53 : 1x9, 1982. 27) Hurd, In vivo and in vitro chemical shift principles, GE NMR, Fremont CA : 28-41. 28) Styles et al., A method of localizing high resolution NMR spectra from human subjects, Mag. res. in medicine, v.2:402- 409, 1985. 29) Mannual- Gyroscan, Phillips. 30)Singh, M., et al., Functional water spectroscopy during visual stimulation using single voxel and ID CSI, SMRM, Second meeting,1994 56

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Three-dimensional functional mapping of the human visual cortex using magnetic resonance imaging

PDF

Seizure activity does not impair LTP induction in young rat hippocampus

PDF

Processing And Visualization In Functional Magnetic Resonance Imaging (Fmri) Of The Human Brain

PDF

Estimation of upper airway dynamics using neck inductive plethysmography

PDF

A multimodal screen reader for the visually impaired

PDF

Auditory brainstem responses (ABR): quality estimation of auditory brainstem responsses by means of various techniques

PDF

Auditory brainstem responses (ABR): variable effects of click polarity on auditory brainstem response, analyses of narrow-band ABR's, explanations

PDF

Application of dynamic mixture model to fMRI

PDF

Propofol Effects On Eeg And Levels Of Sedation

PDF

Respiratory system impedance at the resting breathing frequency range

PDF

A physiologic model of granulopoiesis

PDF

Cross-correlation methods for quantification of nonlinear input-output transformations of enural systems using a Poisson random test input

PDF

Head injury biomechanics: Quantification of head injury measures in rear-end motor vehicle collisions

PDF

An open ear canal sound delivery system

PDF

Cellular kinetic models of the antiviral agent (R)-9-(2-phosphonylmethoxypropyl)adenine (PMPA)

PDF

Ozone resistance of neoprene vulcanizates

PDF

Cross-modal prepulse modification of ERP and startle

PDF

Human Skeletal Muscle Oxygenation And Perfusion: Non-Invasive Measurement By Near-Infrared Spectroscopy

PDF

Proton MR spectroscopy of stimulated auditory cortex

PDF

The development of a process for the extraction of metallic antimony by a hydrometallurgical method

Asset Metadata

Creator Patel, Pankaj B. (author)

Core Title Functional water MR spectroscopy of stimulated visual cortex using single voxel

Degree Master of Science

Degree Program Biomedical Engineering

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag health sciences, radiology,OAI-PMH Harvest

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Singh, Manbir (committee chair), Baudry, Michel (committee member), Khoo, Michael Chee-Kuan (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c20-297340

Unique identifier UC11260262

Identifier EP41387.pdf (filename),usctheses-c20-297340 (legacy record id)

Legacy Identifier EP41387.pdf

Dmrecord 297340

Document Type Thesis

Rights Patel, Pankaj B.

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA