Multi-class motor imagery classification method, device and equipment

The invention discloses a multi-class motor imagery classification method, device and equipment. The method comprises the following steps: acquiring motor imagery data; tracing the motor imagery data according to a tracing model to obtain a source space activation source signal; mapping the source space activation source signal to a plurality of brain regions to obtain a plurality of corresponding brain region signals; constructing a brain region function connection matrix according to the brain region signals; and inputting the brain region function connection matrix into a convolutional neural network for classification to obtain classification result data of multiple types of motor imagery. According to the method, the classification accuracy of four or more types of motor imagery tasks can be improved. 本发明公开了一种多类运动想象分类方法、装置及设备，方法包括获取运动想象数据；根据朔源模型对所述运动想象数据进行朔源，获得源空间激活源信号；将所述源空间激活源信号映射到多个脑区，获得多个对应的脑区信号；根据所述脑区信号构建脑区功能连接矩阵；将所述脑区功能连接矩阵输入至卷积神经网络进行分类，获得多类运动想象分类结果数据。本发明能够提高针对四类及以上类别运动想象任务的分类准确率。