Spike feature extraction algorithms

Two commonly used spike feature extraction algorithms, including PCA [45, 46] and spike peaks [84], are applied to demonstrate the improved sorting results by using the first derivative of spike waveforms. 

Yang Z, Chen T, Liu W (2008) Neuron signature based spike feature extraction algorithm for on- chip implementation. Lecture Proc. 30th Ann. Int. Conf. IEEE EMBS, pp. 1716-1719, August 2008. 

Fig. 22 Feature extraction results using spikes and their derivatives, (a)-(h) display PCA-based feature extraction result using spikes, (i)-(p) display PCA-based feature extraction result using the derivative of spikes (proposed), (q)-(x) display waveform peaks-based feature extraction using spikes, (y)-(af) display waveform peaks-based feature extraction using the derivative of spikes (proposed). All the algorithms are tested without performing interpolation. Overlapping spikes within 1.2 ms are ignored. Two-dimensional features are projected from a higher dimensional space...

Fig. 23 (a) Recorded spikes from cat cerebral cortex are superimposed, (b) the extracted spike features using a subset of samples are plotted and grouped with a clustering algorithm implemented on PC, (c) the classified spike clusters are superimposed, (d)-(k) individual spike clusters superimposed in (c) ate displayed. Spike clusters in (d)-(g) ar plotted in a smaller vertical scale (-0.3,0.15) compared with (b)-(j) in (-0.5,0.3) and (k) in (-0.5,0.5)... 

---