# Removing all R peaks in ECG

5 views (last 30 days)
Ray_Li on 1 May 2020
Answered: Shubh Sahu on 4 May 2020
Hey!
I currently trying to implement this algorithm in MATLAB.
My qeustion is: I need to remove all the R peaks. I have all the Q and S junctions points and I thought about connecting them with interp1 but I doesn't seem to work.
How can I connect all of the Q and S points at once?
Thank you in advance!
This is my code so far: (of course not the best, but it works)
(I have added one ECG file as an example for what I use)
SE = strel('line',2*Fs,0); % 2 seconds ("fast timescale")
SE_1 = strel('line',3*Fs,0); % 3 seconds ("slow timescale")
signal_open = imopen(signal, SE); % Morphologically opens "image"
signal_close = imclose(signal_open, SE_1); % Morphologically closes "opened image"
signal_filtered = signal - signal_close; % Subtracts computed reference value
%% Step 2: R peak detection
%
% Pan-Tompkins Algorithm to detect R peaks and to calculate the RR Intervals
%
[R_Peak_amp,R_Peak_ind] = pan_tompkin(signal_filtered,Fs,0);
if(debug_flag)
%TEST
N = numel(R_Peak_ind);
Max_QRS_duration = 0.12;
Min_QRS_duration = 0.08;
QRS_Duration = (Min_QRS_duration-Max_QRS_duration).*rand(1,N) + Max_QRS_duration;
return;
end
Q_j_ind = zeros(0,length(R_Peak_ind));
S_j_ind = zeros(0,length(R_Peak_ind));
%% Step 3: search for the junctions of Q wave and S wave
first_peak = R_Peak_ind(1);
last_peak = R_Peak_ind(end-1);
q = 1;
s = 1;
while s < param_limit_length_to_N_samples
if(s == first_peak)
R_Peak_ind = R_Peak_ind(2:end);
end
if(s == R_Peak_ind(2))
d = 0;
Left_Peak = R_Peak_ind(1);
Current_Peak = R_Peak_ind(2);
Right_Peak = R_Peak_ind(3);
Y_CurrentPeak = signal_filtered(Current_Peak);
Y_LeftPeak = signal_filtered(Left_Peak);
Y_RightPeak = signal_filtered(Right_Peak);
delta_Y = signal_filtered(Current_Peak) - signal_filtered(Left_Peak);
delta_X = Current_Peak - Left_Peak;
for Pi = Current_Peak:-1:Left_Peak
d = d + 1;
% delta_Y = Y_R - Y_L;
D_Pi = abs(delta_Y * Pi - delta_X * signal_filtered(Pi) + Current_Peak * Y_LeftPeak - Y_CurrentPeak * Left_Peak)/sqrt(delta_Y^2 + delta_X^2);
W_Pi = cos(2*pi*(Pi-Current_Peak)/(Current_Peak-Left_Peak));
WD_Pi_q = D_Pi * W_Pi;
WD_Pi_amp_q(d) = WD_Pi_q;
end
half = floor((Current_Peak-Left_Peak)/2);
[~,Q_j] = max(WD_Pi_amp_q(1:half));
Q_j_ind(q) = Current_Peak - Q_j + 1;
delta_Y = signal_filtered(Right_Peak) - signal_filtered(Current_Peak);
delta_X = Right_Peak - Current_Peak;
d = 0;
for Pi = Current_Peak:Right_Peak
d = d + 1;
% delta_Y = Y_R - Y_L;
D_Pi = abs(delta_Y * Pi - delta_X * signal_filtered(Pi) + Right_Peak * Y_CurrentPeak - Y_RightPeak * Current_Peak)/sqrt(delta_Y^2 + delta_X^2);
W_Pi = cos(2*pi*(Pi-Current_Peak)/(Right_Peak-Current_Peak));
WD_Pi_s = D_Pi * W_Pi;
WD_Pi_amp_s(d) = WD_Pi_s;
end
half = floor((Right_Peak-Current_Peak)/2);
[~,S_j] = max(WD_Pi_amp_s(1:half));
S_j_ind(q) = Current_Peak + S_j - 1;
q = q + 1;
R_Peak_ind = R_Peak_ind(2:end);
end
if(s == last_peak)
break;
end
s = s+1;
end
subplot(2,1,1)
plot(t,signal_filtered,'-o','MarkerEdgeColor','b','MarkerIndices',Q_j_ind);
subplot(2,1,2)
plot(t,signal_filtered,'-o','MarkerEdgeColor','r','MarkerIndices',S_j_ind);

Shubh Sahu on 4 May 2020
As you have mentioned that you have all Q and S points. I assume that you must have respective sample time for Q and S. Use interp1 between Q-S segment and replace the R information from at those time sample. For more information please go through following link:

### Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by