Fig. 1
From: Spike pattern recognition by supervised classification in low dimensional embedding space
Spike detection framework. The 1st step of the method detects spike-like waveforms by extracting the two half-waves. The half-waves are defined between the negative peak (marked with a red circle) and the two positive peaks (marked with green stars) and are characterized by the amplitude difference (A 1, A 2) and duration (D 1, D 2). The 2nd step of the method classifies the detected transients into spikes and non-spikes using machine learning techniques. (Color figure online)