fMRI temporal clustering analysis in patients with frequent interictal epileptiform discharges: Comparison with EEG-driven analysis
fMRI temporal clustering analysis in patients with frequent interictal epileptiform discharges: Comparison with EEG-driven analysis
复制标题
DOI：
10.1016/j.neuroimage.2005.01.052
复制
发表时间：
2005-05-15
期刊：
影响因子：
5.7
通讯作者：
Lemieux, L
中科院分区：
文献类型：
--
作者：
Hamandi, K;Haddadi, AS;Lemieux, L
时间聚类分析（TCA）是一种探索性的数据驱动技术，已被提出用于分析静息功能磁共振成像定位癫痫样活动，而不需要同时脑电图。传统上，癫痫活动的fMRI仅限于那些具有细微临床事件或频繁发作间期癫痫样EEG放电的患者，需要同时进行EEG记录，从中导出线性模型以从fMRI数据进行有效的统计推断。我们试图评估TCA的结果与EEG相关的功能磁共振成像在8个选定的情况下进行比较。根据一致的EEG和结构MRI结果，以及EEG衍生的fMRI分析中观察到的一致激活，选择具有明确致癫痫定位或定侧的病例。在三个中，用TCA观察到激活区域，但没有对应于用EEG驱动分析获得的电临床定位或激活。时间簇与最大头部运动的时间密切一致。我们认为这是对这种方法的严重混淆，并建议谨慎对待不涉及运动和生理噪声的TCA解释。新的技术，以定位癫痫活动与功能磁共振成像单独需要验证与适当的独立措施。在发作间期癫痫样活动的研究中，最好采用同步EEG记录。(c)2005年爱思唯尔公司All rights reserved.
Temporal clustering analysis (TCA) is an exploratory data-driven technique that has been proposed for the analysis of resting fMRI to localise epileptiform activity without need for simultaneous EEG. Conventionally, fMRI of epileptic activity has been limited to those patients with subtle clinical events or frequent interictal epileptiform EEG discharges, requiring simultaneous EEG recording, from which a linear model is derived to make valid statistical inferences from the fMRI data. We sought to evaluate TCA by comparing the results with those of EEG correlated fMRI in eight selected cases. Cases were selected with clear epileptogenic localisation or lateralisation on the basis of concordant EEG and structural MRI findings, in addition to concordant activations seen on EEG-derived fMRI analyses. In three, areas of activation were seen with TCA but none corresponding to the electro-clinical localisation or activations obtained with EEG driven analysis. Temporal clusters were closely coincident with times of maximal head motion. We feel this is a serious confound to this approach and recommend that interpretation of TCA that does not address motion and physiological noise be treated with caution. New techniques to localise epileptogenic activity with fMRI alone require validation with an appropriate independent measure. In the investigation of interictal epileptiform activity, this is best done with simultaneous EEG recording. (c) 2005 Elsevier Inc. All rights reserved.