Automated segmentation of neural recordings for optimal on-line recognition of neural waveforms

We present an iterative algorithm for separating the segments containing exclusively neural noise in extracellular recordings without prior knowledge of neural spike locations or waveforms. This allows on-line design of a whitening filter and on-line determination of thresholds for detection and classification of neural spikes without human supervision. This algorithm can also be used as a first data reduction phase for the detection task.