Spike detection algorithm improvement, spike waveforms projections with PCA and hierarchical classification

Definition of single spikes from multiunit spike trains plays a critical role in neurophysiology and in neuroengineering. Moreover, long period analysis are needed to study synaptic plasticity effects and observe the long and medium term development on which all central nervous system (CNS) learning functions are based. Therefore, the increasing importance of long period recordings makes necessary on-line and real time analysis, memory use optimization and data transmission rate improvement. A threshold-amplitude spikes detection method is chosen and 5 noise level estimate methods were developed. Than APs are bundled to group using principal component analysis and classified (hierarchical classifier). The system has lot of applications like high-throughput pharmacological screening and monitoring effects.