Double Simulated Annealing for Functional MRI Analysis

使用雙重模擬退火法於腦部功能性磁振造影影像分析之研究

葉進儀(Jinn-Yi Yeh) ； 傅家啓(Ja-Chih Fu)

《工業工程學刊》 22卷6期 (2005/11) Pp. 497-508

https://doi.org/10.29977/JCIIE.200511.0006

摘要

本研究使用雙重模擬退火法，針對功能性磁振造影產生的影像加以分析，第一個模擬退火法先去除頭蓋骨以外的部分，剩下之影像再由第二個模擬退火法找出活化區的位置，而績效評估方式採用繪製ROC(Receiver Operating Characteristic)曲線、相似度分析與比較其他分析的方法，例如兩樣本t檢定法、二個數列的交互相關係數及一般的線性組合等方法，實驗結果發現，雙重模擬退火法比其他方法更能正確地分辨出功能性磁振造影之活化區。

關鍵字

功能性磁振造影影像 ； 模擬退火法 ； ROC曲線 ； 相似度分析 ； 活化區

並列摘要

This report proposes a double simulated annealing (DSA) for functional magnetic resonance image (fMRI) analysis. The first simulated annealing (SA) is used to disconnect the brain from the skull. The second SA is applied to locate the activation area of the fMRIs. The performance evaluation of this approach includes receiver-operating characteristic (ROC) analysis, similarity analysis, and comparisons with other analytical methods such as classical SA (CSA), a t-Test, cross-correlation (CCR), and a general linear model (GLM). Experimental results show that the proposed algorithm can get better solutions than other approaches.

並列關鍵字

functional magnetic resonance imaging ； simulated annealing ； receiver-operating characteristic ； similarity analysis ； activation area

參考文獻

Baker, J.R.,R.M. Weisskoff,C.E. Stern(1994).Statistical assessment of functional MRI signal change.(Proceedings of the 2nd Annual Meeting of the Society of Magnetic Resonance).

Bandettini, P.A.,A. Jesmanowicz,E.C. Wong,J.S. Hyde(1993).Processing strategies for time-course data sets in functional MRI of the human brain.Magnetic Resonance in Medicine.30,161-173.

Baumgartner, R.,C. Windischberger,E. Moser(1998).Quantification in functional magnetic resonance imaging: fuzzy clustering vs. correlation analysis.Magnetic Resonance Imaging.16,115-125.

Baumgartner, R.,L. Ryner,W. Richter,R. Summers,M. Jarmasz,R. Somorjai(2000).Comparison of two exploratory data analytical methods for fMRI: fuzzy clustering vs. principal component analysis.Magnetic Resonance Imaging.18,89-94.

Baumgartner, R.,R. Somorjai,R. Summers,W. Richter(2001).Ranking fMRI time courses by minimum spanning trees: assessing coactivation in fMRI.NeuroImage.13,734-742.