Development of intelligent system for classification of multiple human brain states corresponding to different real and imaginary movements
In this paper we considered the technique for classification of different types of electroencephalogram activity corresponded to different types of real and imaginary movement. We developed experimental design and performed series of experiments on volunteers for real and imaginary movements of hands and legs. We proposed a system based on artificial neural networks for classification of multiple human brain states corresponding to different real and imaginary movements. We provided testing and found optimal for this task size of input signal and structure of artificial neural network.
CYBERNETICS AND PHYSICS, Vol. 6, No. 3. 2017, 103-107