How can ı do spike sorting to filtered EEG signal.

I have this sample and ı need to do spike sorting. I start with finding thresholding value and want to see changes with this code. I have trouble with plotting in for loop and also further steps. I am very new in Matlab. If you can help it would be great thank you.

for k = 1:0.1:5
    
omg = median(abs(samp)/0.6745);
Thr =k * omg;
[pks, locs] = findpeaks(samp);
findpeaks(samp, 'Threshold', Thr);
[r,NumPeaks] = size(pks);
NumPeaks

end
hold on

Answer 1

[Star Strider]

Try this:

D = load('samp.mat');
samp = D.samp;
data = 1E-3;                                                    % What Are ‘data’?
omg = median(abs(data)/0.6745);
% th = 1:0.1:5;
th = 5 : -0.1 : 1;                                              % Easier To Count Down Here
for k = 1:numel(th)
    Thr = th(k) * omg;
    [pks{k}, locs{k}] = findpeaks(samp, 'Threshold', Thr);      % Peaks Above This ‘Thr’ Value
    kprv = max(k-1,1);                                          % Previous Index Value
    lv = ~ismember(locs{k},locs{kprv});                         % Spikes Above Current ‘Thr’ & Below Previous ‘Thr’
    nspikes(k) = nnz(lv);                                       % Number Of Spikes Above Current ‘Thr’ & Below Previous ‘Thr’
    ThrPks{k} = pks{k}(lv);                                     % Peak Amplitudes Above Current ‘Thr’ & Below Previous ‘Thr’
    ThrLcs{k} = locs{k}(lv);                                    % Locations Of Peak Amplitudes Above Current ‘Thr’ & Below Previous ‘Thr’
end

figure
bar(th, nspikes)
xlabel('Threshold')
ylabel('# Spikes')

I am not certain what you want, so I saved everything about your data in my code.

[kaan bakircioglu]

Thank you very much for this code. I needed this because ı need to set an empirical rule for the ? parameter so that the number of spikes for each extracellular signal is close to the real number.

[Star Strider]

My pleasure.

If my Answer helps you solve your problem, please Accept it!

Answer 2

[Image Analyst]

Not sure what you want to do. Do you just want to order them based on height at the peaks? If so

[sortedPeakValues, sortOrder] = sort(pks);
sortedIndexes = loc(sortOrder);

[kaan bakircioglu]

Actually ı want to do make the figure #spikes vs. ? . Thank you for you answer.

[Image Analyst]

Try this:

clc;    % Clear the command window.
close all;  % Close all figures (except those of imtool.)
clear;  % Erase all existing variables. Or clearvars if you want.
workspace;  % Make sure the workspace panel is showing.
format long g;
format compact;
fontSize = 15;
s = load('samp.mat')
samp = s.samp;
factor = 1:0.1:5	
for k = 1 : length(factor)
	omg = median(abs(samp) / 0.6745);
	Thr = factor(k) * omg;
	[pks, locs] = findpeaks(samp, 'Threshold', Thr);
	numberOfPeaks(k) = length(pks);
end
plot(numberOfPeaks, 'b*-', 'LineWidth', 2);
grid on;
xlabel('k', 'FontSize', fontSize);
ylabel('Number Of Peaks', 'FontSize', fontSize);
yticks(0:15);
hold on

[kaan bakircioglu]

Thank you for your effort do you have any suggestion about how can ı set an empirical rule for the ? parameter so that the number of spikes for each extracellular signal is close to the real number.

[Image Analyst]

It's really the threshold that you want to set, not the k. I'm not sure which peaks in the data are desired one and which are noise ones. The data itself doesn't say - basically it's a judgment call. There is no one, true, God-given correct answer. So you have to specify the threshold using your expert knowledge. Whatever works best for you is the right answer.

[kaan bakircioglu]

Yes you are right, thank you!