Neuron firing spike signal picking method, device and computer equipment

Abstract

The present application relates to a method, a device, a computer device and a storage medium for picking out neuron discharge spike signals. The method includes: preprocessing the discharge signal of the neuron to be picked to obtain a continuous filtered discharge signal; performing energy calculation on the continuously filtered discharge signal to obtain a continuous energy signal; performing peak statistics on the continuous energy signal to determine the energy peak value; Establish a cumulative distribution function of chi-square distribution according to the continuous energy signal for analysis to determine the picking threshold; filter the energy peak through the preliminary peak extraction model constructed by the picking threshold to obtain a preliminary peak; Extracting signal segments from the continuously filtered discharge signal to obtain a discharge signal segment; analyzing the discharge signal segment according to the picking threshold to determine a neuron discharge spike signal. Through this method, the robustness of picking neuron firing spikes is stronger, and the accuracy of picking results is also improved.

Description

technical field [0001] The present application relates to the technical field of digital signal processing, in particular to a method, device, computer equipment and storage medium for picking out neuron discharge spike signals. Background technique [0002] When collecting neural signals, a high sampling rate method is used to record the continuous signals around the neurons, and the neuron spike signal (Spike) is a short discharge process, lasting about 1ms. In order to extract this short-term signal, the commonly used method is to observe the signal, set a threshold, and select the signal that exceeds the threshold as a firing process of a neuron. [0003] This method requires people to observe the signal threshold, which increases the workload, and because the threshold is artificially specified, it increases the instability of signal extraction; and this method can only pick out the neural discharge spikes with large amplitudes. However, some spikes with smaller amplit...

AI Technical Summary

Problems solved by technology

[0003] This method requires people to observe the signal threshold, which increases the workload, and because the threshold is artificially specified, it increases the instability of signal extraction; and this method can only pick out the neural discharge spikes with large amplitudes. However, some spikes with smaller amplitudes will be ignored in this method, resulting in statistical errors in neural signals.

[0004] Therefore, the accuracy of neuron firing spikes picked by the current neural firing spike picking method is low

Images