基于脑电复杂度的意识任务的特征提取与分类

本研究提出了利用事件相关脑电复杂度提取大脑运动意识特征，应用Mahalanobis距离判别式分析法，对人脑想象左右手运动任务进行分类，获得了满意的结果。对受试者想象左右手运动期间在大脑初级感觉运动皮层区记录的脑电信号采用复杂度分析方法量化了事件相关去同步（ERD）和事件相关同步（ERS）时程，结果表明EEG复杂度特征较好反映了ERD／ERS变化时程。最后对测试数据进行分类，最大分类正确率达到86．43％，通过最大分类正确率，最大信噪比，最大互信息等评价指标比较，验证了该方法的有效性，从而为大脑运动意识任务的特征提取及分类提供了新思路。

Feature Extraction and Classification of Consciousness Task Based on EEG Complexity

The method extracting brain motor consciousness feature by event-related EEG complexity was proposed in this paper. Applying Mahalanobis distance discriminant analysis to distinguish the left and right hand motor imaginary tasks,the satisfactory results were obtained, The data from BCI competition 2003 provided by Graz University of Technology were analyzed. ERD/ERS time course was quantified by complexity analysis method. The results showed that EEG complexity feature effectively reflected ERD/ERS time course changes. Finally, the test data were analyzed and the analysis results were evaluated from three performances including the maximum classification accuracy, maximum SNR, maximum mutual information. The evaluation results indicated that the method presented in this paper was effective and it might provide a new approach to identify brain consciousness tasks.