Optimal discrimination and classification of neuronal action potential waveforms from multiunit, multichannel recordings using software-based linear filters

Abstract
We describe advanced protocols for the discrimination and classification of neuronal spike waveforms within multichannel electrophysiological recordings. The programs are capable of detecting and classifying the spikes from multiple, simultaneously active neurons, even in situations where there is a high degree of spike waveform superposition on the recording channels. The protocols are based on the derivation of an optimal linear filter for each individual neuron. Each filter is tuned to selectively respond to the spike waveform generated by the corresponding neuron, and to attenuate noise and the spike waveforms from all other neurons. The protocol is essentially an extension of earlier work [1], [13], [18]. However, the protocols extend the power and utility of the original implementations in two significant respects. First, a general single-pass automatic template estimation algorithm was derived and implemented. Second, the filters were implemented within a software environment providing a greatly enhanced functional organization and user interface. The utility of the analysis approach was demonstrated on samples of multiunit electrophysiological recordings from the cricket abdominal nerve cord.