An artificial neural network approach to predicting arm movements from ECoG

There are three specific aims. First, demonstrate the practicality of using an artificial neural network based approach to correlate these cortical signals with actual and imagined arm movements. Second, to identify areas of the cortical surface that provide the most useful command information. Third, quantify the information content and information transfer rate of the signals obtained from the subdural grids relative to a set of relevant arm movements. This work presents progress toward these aims.