时间聚类分析功能MRI对癫痫病灶定位的初步研究

A preliminary functional MRI study with temporal clustering analysis for localization of epileptic activity

目的 探讨癫痫病灶的定位方法。方法对12例颅内有明确局灶性病变的癫痫患者行静息状态下的功能MR成像（fMRI）。采用时间聚类分析（TCA）软件程序实现TCA算法，并对fMRI数据进行分析，经统计分析后得到脑解剖-功能融合图像，观察病变区有无激活反应，并对TCA反应聚类时间曲线与统计参数图（SPM）头部运动曲线进行比较。结果观察脑解剖-功能融合图像发现，12例中8例在病变区有激活反应，4例未发现明显病灶区激活反应。病灶区无论有无激活反应，反应幅度较大的TCA反应聚类时间点和头部运动曲线之间有一定的对应关系。结论TCA作为一种探索性数据驱动方法，一定程度上可以对癫痫的发作间期痫样放电活动进行检测和定位，但其结果会受到头部运动的影响，分析时需考虑到生理、病理及运动等多种因素。

Objective To study the potential of using temporal clustering analysis （TCA） technique in localizing an epileptogenic zone. Methods Twelve patients with epilepsy were examined using resting functional MRI （fMRI）. The patients had detectable focal lesions on cranial MRI. TCA was performed to analyze resting fMRI data in order to identify the timing of interictal epileptiform discharges （IEDs）. Standard event-related fMRI analysis in SPM99 was used to generate maps of the activation induced by epileptic brain activities. Comparisons were made between TCA results and SPM motion trochoid. Results Eight of the twelve subjects showed activations in the brain regions that were consistent with those lesions determined on anatomic MRI. The remaining four subjects showed no clear activation in the areas of detectable lesions. In addition, correlation was found between TCA results and motion trochoids. Conclusions Temporal cluster analysis, an exploratory data-driven technique, may provide the timing information about interictal epileptiform discharges. However, the results from this novel fMRI analytical technique need to be interpreted with caution as it is vulnerable to motion artifact.