Detection and sorting of neural spikes using wavelet packets

We propose a novel method for the detection and sorting of recorded neural spikes using wavelet packets. We employ the best basis via the Shannon's in formation cost function and local discriminant basis using mutual informati on. We demonstrate the efficiency of the method on data recorded in vitro f rom 2D neural networks. We show that our method is superior both in separat ion from noise and in identifying superimposed spikes.