Analysis of Neural Data

This repository includes useful MATLAB codes for ENG analysis.

There are three main stages in the algorithm: (1) spike detection, (2) spike feature selection (Wavelet), and (3) clustering of the selected spike features (K-means).

In the first step, spikes are detected with an automatic amplitude threshold on the band-pass filtered data. In the second step, a small set of wavelet coefficients from each spike is chosen as input for the clustering algorithm. Finally, the K-means classifies the spikes according to the selected set of wavelet coefficients.

• Fs = 30000 Hz; % Sampling frequency
• F_low = 300 Hz; % low pass filter for detection
• F_high =3000 Hz; % high pass filter for spike detection

Threshold:

• T_min = 5; % minimum threshold for estimated noise
• T_max = 12; % maximum threshold for avoid high amplitude artifact

Detect spike times:

• w_pre = 20; % w_pre datapoints before the spike peak are stored
• w_post =40; % w_post datapoints after the spike peak are stored

Please Run Code Main.m, the Software will run automatically and there is no need to run test code

[email protected]