林發暄Lin, Fa-Hsuan臺灣大學:醫學工程學研究所周育震Chou, Yu-ChenYu-ChenChou2010-05-182018-06-292010-05-182018-06-292009U0001-2907200914395400http://ntur.lib.ntu.edu.tw//handle/246246/183729在這篇研究中，我們以結構方程式模型(Structural Equation Modeling, SEM)來估計由功能性磁振造影(functional magnetic resonance imaging, fMRI)記錄下的人腦內各個活動區域之間的交互作用，藉此來研究人腦受到刺激時不同區域之間的有效連結(effective connectivity)。由於功能性磁振造影訊噪比(signal-to-noise ratio, SNR)不高，因此本論文專注研究雜訊對SEM估計結果的影響。首先我們使用數值模擬的方法，來找出結構方程式模型不同區域之間的路徑係數(path coefficients)在受到不同訊噪比所造成的影響。另外，由於功能性磁振造影在時間解析度上較差，需要較長的時間來取樣，所以我們每回實驗所能取得的時間序列長度有限。因此在資料長度不同的情況下，我們估計訊號長度對結構方程式模型的路徑係數分布情況所產生的影響。在本篇論文中，我們也將結構方程式模型運用到兩種功能性磁振造影影像實驗(視覺認知的實驗以及針灸刺激實驗)上。這篇研究中的分析方法可以提供資訊讓我們對已知的結構方程式模型內部區域連結進行修改，並且對計算出的連結係數提供統計推論。這種方法也能用在其它人腦功能性影像資料上。This study aims at optimizing Structural Equation Modeling (SEM) analysis in order to accurately estimate the effective connectivity between active brain regions elucidated by the functional magnetic resonance imaging (fMRI) of the human brain during tasks and cognition. Since the empirical fMRI data are of relatively low signal-to-noise ratio (SNR), we are interested in the effects of noises over SEM estimates. First, we use numerical simulation to evaluate the SNR sensitivity of the estimated path coefficients in the SEM. In addition, we also investigate the dependency of path coefficients on the number of data samples, which are practically limited by the relative slow sampling rate (~2 second per volume). Based on the estimated distributions of path coefficients, we quantify the variability of the path coefficients when SNRs and data lengths vary. In this thesis, we also apply the SEM to respective in vivo fMRI experiments to study causal modulations among brain areas during visual cognition and acupuncture stimulus. The SEM analysis developed in this thesis can suggest likely modifications of the a priori directional connectivity required in the traditional SEM and offers statistical inferences on the path coefficients. This method can be used for other brain imaging data.LIST OF CONTENTS IIIST OF FIGURES VIIST OF TABLES IX文摘要 XIBSTRACT XII. INTRODUCTION 1. MATERIAL AND METHODS 3.1. COVARIANCE MATRIX 3.2. STRUCTURAL EQUATION MODELING 4.2.1. Maximum likelihood estimation 8.2.2. Modification index 10.3. STATISTICAL TESTS 12.3.1. Chi-square test 12.3.2. One-sample t-test 13.3.3. Paired t-test 13. SIMULATIONS 15.1. INTRODUCTION 15.2. CONSTRUCT AN EFFECTIVE CONNECTIVITY MODEL 16.3. SIMULATE THE TIME SERIES 17.4. SIMULATE THE NOISE 19.5. SIMULATION PARAMETER 20.6. RESULTS AND DISCUSSIONS 21.6.1. Normal distribution of estimated path coefficients 21.6.2. Effect of data length 22.6.3. Effect of noise 25. APPLICATION OF THE SEM TO FMRI 31.1. INTRODUCTION 31.1.1. SPM preprocessing 31.1.2. Realignment 32.1.3. Slice timing correction 32.1.4. Coregisteration 33.1.5. Normalization 34.2. GENERAL LINEAR MODEL 35.3. LIKELIHOOD RATIO TEST 38.4. VISUAL COGNITION EXPERIMENT 40.4.1. Stimulus, task, and data acquisition parameters 40.4.2. Mapping of active brain areas 42.4.3. SEM analysis 45.4.4. Results 46.4.4.1. Original Model 46.4.4.2. The model with added path R10/L46 to B22 52.4.5. Discussion 57.5. EXPERIMENT OF ACUPUNCTURE STIMULATION 61.5.1. Stimulation and data preprocessing 61.5.2. Results 63.5.2.1. SEM result using time series without GLM processing 63.5.2.2. SEM result using time series with GLM processing 68.5.3. Discussion 73. CONCLUSION 76LOSSARY 77EFERENCES 79application/pdf1793171 bytesapplication/pdfen-US結構方程式模型功能性磁振造影有效連結路徑係數視覺認知Structural Equation ModelingSEMfunctional Resonance MagneticImagingfMRIeffective connectivitypath coefficientvisual cognitionthesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/183729/1/ntu-98-R96548054-1.pdf