在Resting-state fMRI下以Seed-Based分析評估不同磁場、掃描時間和Seed ROI半徑造成的功能性連結變化

Empirical Evaluations of Magnetic Field Strength, Duration and Region of Interest (ROI) Radius Effects in Seed-Based, Resting-State Functional Magnetic Resonance Imaging Analyses

謝政學(Zheng-Xue Xie)

指導教授 ： 陳佳伶

中山醫學大學/醫學科技學院/生物醫學科學學系碩士班/碩士(2012年)

https://doi.org/10.6834/CSMU.2012.00117

摘要

最近幾年休息狀態fMRI實驗研究證明神經網路可能因外在環境、呼吸和心跳生理雜訊影響以及此研究在極低頻(0.01~0.08Hz)狀態下觀察，因此功能性神經連結結果隨時間可能會有些許改變。本研究主要目的，使用種子相關分析(seed-based)方法研究休息狀態fMRI，探討磁場強弱、掃描時間長短以及感興趣種子區域半徑大小可能造成的功能性連結和相關係數的變化情形，預期找出在1.5T和3T MRI下主要的5個休息狀態功能性連結網路其可能理想的掃描時間和感興趣種子區域半徑範圍。願在未來以種子相關分析的休息狀態fMRI研究或者在臨床上使用，提供一個理想參數建議。
本實驗共有19位受測者〈9位1.5T和10位3T〉，每位皆接受30分鐘fMRI休息狀態掃描。(1)掃描時間研究中，以200秒間隔，將30分鐘數據複製並截斷產生200秒到1600秒不同時間數據。(2)感興趣種子區域半徑研究，以3mm半徑為間隔，分析3mm到30mm共10組不同半徑對功能性連結影響。3個變數研究皆以種子相關分析。
我們得到以下幾個實驗結果：(a)臨床最短掃描時間設定，我們建議在3T MRI五個神經網路可能最少皆需要約200秒(3分)左右掃描時間；而在1.5T MRI五個神經網路需要的掃描時間是不同的，我們建議運動神經網路最少需400秒(7分)左右、預設模式和視覺神經網路最少需200秒(3分)左右以及在邊緣系統的杏仁核和海馬迴神經網路需要掃描時間較長約800秒(13分)左右。(2)感興趣種子區域半徑研究，我們發現感興趣種子區域半徑增大確實是會讓功能性連結的網路區域增加以及較高的相關係數。但我們發現一昧增大感興趣種子區域半徑是錯誤的，因為考慮到感興趣種子區域之腦區大小例如邊緣系統本身腦區不大可能造成圈選區域錯誤，進而造成失真和錯的功能性連結，所以我們建議理想seed ROI半徑範圍是3mm到9mm。
臨床上建議使用3T MRI做休息狀態fMRI的掃描，其影像品質較好、掃描時間短以及病人的忍受度也比較高，對於小朋友、危急和不自主亂動的病人來說是很好的。但如果醫院沒有3T MRI也沒關係，1.5T MRI可找到正確的功能性連結位置，再經由後處理適度調整感興趣種子區域半徑，影像品質是可以接受的。

關鍵字

休息狀態功能性磁振造影 ； 功能性連結 ； Seed-based分析

並列摘要

Resting-state functional magnetic resonance imaging (rs-fMRI) is widely used for investigating spontaneous brain activities and subsequently revealing functional connectivity. Resting-state brain connectivity has been proven to provide insight into the neuropathophysiology of clinical condition. Previous studies have shown that different preprocessing and data acquisition parameters that vary across laboratories account for the variations of connectivity maps in terms of the strength and spatial extent. However, the effects of remaining procedures were rarely investigated and the significance of different data acquisition parameters and analysis steps on rs-fMRI need to be studied. Therefore, the main purpose of this study is to evaluate the effects of different magnetic field strength and scan duration during acquisition steps, and the size of seed selection in analysis procedures to provide recommendations for optimization that can be useful in clinical rs-fMRI applications.
Nineteen volunteers were recruited for the rs-fMRI study. The experiments were conducted on 1.5 T and 3.0T scanners, respectively. To study the optimal scan duration, a total of 900 measurements were scanned using TR=2s, corresponding to 30 minutes of total acquisition time. 8 sets of data with different scan duration ranging from 200 to 1600 sec with 200 sec increment were then grouped for a further seed-based correlation analysis on 5 brain networks (i.e. motor, default, visual, amygdala, and hippocampus networks). The effect of seed size on rs-fMRI was evaluated using 10 various sizes of seed with the radius ranged from 3 to 30 mm with 3 mm increment.
Overall, duration of about 200 sec (3 minutes) of resting state data acquisition was found to provide a good compromise between total experiment time and robust functional connectivity in both 1.5 T and 3.0 T rs-fMRI studies. However, due to the nature of lower signal-to-noise ratio of 1.5 T MRI, a longer scans (up to 800 sec, 13 min) are preferable especially for networks like amygadala and hippocampus. Meanwhile, the size of seed selection had shown the effects on the strength and the spatial extent of the connectivity maps and the radius of seed size ranging from 3 to 9mm were found to produce every network robustly.
In conclusion, duration of 3 minutes of 3T resting state data acquisition is preferable in rs-fMRI studies; however, to provide better rs-fMRI results in 1.5T MRI if access to 3T MRI is impossible, using a slightly longer scan duration depending on the networks of interest and selecting seed size of 3 mm to 9 mm radius is recommended.

並列關鍵字

Rest-state fMRI ； functional connectivity ； seed-based analyses

參考文獻

2. Raichle, M.E., Circulatory and Metabolic Correlates of Brain Function in Normal Humans. Comprehensive Physiology, 1987. 5.

3. Ogawa, S., et al., Brain magnetic resonance imaging with contrast dependent on blood oxygenation. Proc Natl Acad Sci U S A, 1990. 87(24): p. 9868-72.

4. Ogawa, S., et al., Intrinsic signal changes accompanying sensory stimulation: functional brain mapping with magnetic resonance imaging. Proc Natl Acad Sci U S A, 1992. 89(13): p. 5951-5.

5. Turner, R., et al., Echo-planar time course MRI of cat brain oxygenation changes. Magn Reson Med, 1991. 22(1): p. 159-66.

6. Ogawa, S., et al., Oxygenation-sensitive contrast in magnetic resonance image of rodent brain at high magnetic fields. Magn Reson Med, 1990. 14(1): p. 68-78.