Zhang Jing, Buchsbaum Monte S, Chu King-Wai, Hazlett Erin A
Department of Bioinformatics School of Biomedical Engineering, Capital Medical University, Beijing 100069, P.R. China.
Stud Health Technol Inform. 2013;192:921.
This study is a method comparison of three region-of-interest (ROI) tracing approaches. fMRI images acquired during an event-related frequency discrimination task were used in this study. Three [(1) manual, (2) semi-automated, (3) automated] approaches were applied to generate the ROI in the caudate nucleus and extract fMRI hemodynamic response (HR) from the ROI. Less activation with weak and delayed fMRI HR in the caudate were observed in patients than controls by the 3 approaches, but the manual and automated approaches were more sensitive in detecting such fMRI abnormalities. High correlations of the HR measures were found between the manual and semi-automated ROI approaches, suggesting that the semi-automated approach could be used as an alternative to the manual approach in detecting fMRI experimental effect.
本研究是对三种感兴趣区域（ROI）追踪方法的比较。本研究使用了在事件相关频率辨别任务期间采集的功能磁共振成像（fMRI）图像。应用三种方法（(1)手动、(2)半自动、(3)自动）在尾状核中生成ROI，并从ROI中提取fMRI血流动力学反应（HR）。通过这三种方法观察到，与对照组相比，患者尾状核中fMRI HR较弱且延迟，激活较少，但手动和自动方法在检测此类fMRI异常方面更敏感。手动和半自动ROI方法之间的HR测量值具有高度相关性，这表明半自动方法在检测fMRI实验效果方面可作为手动方法的替代方法。
