Abstract
Thresholding methods usually detect neural spikes or bursts of electrical pulses in a cortical electrophysiological signal. Due to the possibility of multiple threshold crossings representing the same spike, it is essential to set spike temporal boundaries to prevent double detection appropriately. We consider three spike validation methods and compare their perfor- mance on a set of simulated electrophysiological data mimicking the spiking activity of a neuron with different amounts of background noise. Since the results of one method rapidly deteriorated with decreasing signal-to-noise ratio, we formulate a modification leading to the algorithm’s performance improvement.
Go back