How do I find location of my specific peak in ECG signal?

I'm trying to find the heartbeats from my ECG signal by by finding the locations of peaks. I have written the code for butterworth filter and had to invert my signal to find signal since my T wave and R wave were the same lenght. So I'm stuck trying to write a code that passes through y= 0.10 which would hit all the inverted signal and get the location each x that coresponds to that peak. Here is what I have for my end result :

sig_filt = -(sig_filt) %inverse

figure

subplot(2,1,1)

plot(T,-(sig_filt));

xlim([0,60])

subplot(2,1,2)

plot(T,-(sig_filt));

xlim([0,5])

subplot(2,1,1)

plot(T,-(sig_filt));

xlim([0,60])

subplot(2,1,2)

plot(T,-(sig_filt));

xlim([0,5])

findpeaks(-sig_filt);

[pks,min_locs] = findpeaks(-(sig_filt),"MinPeakHeight",0.11);

Answer 1

[Star Strider]

‘I have written the code for butterworth filter and had to invert my signal to find signal since my T wave and R wave were the same lenght.’

I hope you mean ‘height’ rather than ‘length’, since the QRS duration is ordinarly less than 120 ms while the T duration is typically about 200 ms. With this clarification, use the 'MinPeakDistance' name-value pair to force findpeaks to ignore the T deflection, since it will always be after the R deflection. (There is typically no need to invert the EKG to use findpeaks, unless the intent is to locate Q deflections.)

EDIT — (18 Jul 2022 at 15:52)

There is baseline drift (variation) that complicates things. Remove it first, then use 'MinPeakProminence' to isolate the R deflections —

F = openfig('ecg.fig');

sp211 = subplot(2,1,1);

Lines = findobj(sp211, 'Type','line');

X = Lines.XData; % Recover Data

Y = -Lines.YData; % Recover Data & Invert

Fs = 1/(X(2)-X(1)); % Sampling Frequency

Yf = highpass(Y,2,Fs, 'ImpulseResponse','iir'); % Remove Baseline Drift

[Rwa,locs] = findpeaks(Yf, 'MinPeakProminence',0.25); % Detect R-Deflections

figure

plot(X, Yf)

hold on

plot(X(locs), Rwa, '^r')

hold off

grid

title('Original')

figure

plot(X, Yf)

hold on

plot(X(locs), Rwa, '^r')

hold off

grid

xlim([10 15])

title('Zoomed')

This EKG displays significant pathology likely due to ischemia, with prominent S-T depression, notched P deflections suggestive of an intraatrial conduction delay, and prominent T deflections. I would have to see the full 12-lead version to venture a diagnosis. (This does not appear to be a Lead II EKG, and may be a Lead III.)

.

[Niloofar Roshanfekr]

I guess I was using height and length interchanchably, sorry. I have attached the plot of the inverted signal and you can see the height of the R wave and the T wave are very much the same. so if I were to use the not inverted data, I would not be able to distingush between the two peaks.

[Niloofar Roshanfekr]

Thank you, the purpuse of this project is to write a code to calculate the heart beat, so the data collection was only done by 3 electrodes since the focus was not on diagnosis.

[Star Strider]

As always, my pleasure!