How to create a wiener filter?

Hello,

I have a problem creating a wiener filter. The purpose of this exercise was to design a wiener filter and apply it to a given ecg signal.

I have created a variable 'model' which contains the desired optimal signal, I created this with ones and linspace, visually.

Then I selected 4 T-P intervals, which are iso-electric and thus can be used for modelling the sound. (I plotted the signal and used ginput to select the points)

%P-T-segments

A1= [ecg(x1(1):x1(2))]; A2= [ecg(x1(3):x1(4))]; A3= [ecg(x1(5):x1(6))]; A4= [ecg(x1(7):x1(8))];

%removing DC-component by substracting mean

A1=A1-mean(A1); A2=A2-mean(A2); A3=A3-mean(A3); A4=A4-mean(A4);

%calculating PSD of noise segemnts

noise = zeros([4,700]); noise(1,:) = abs(fft(A1,700)).^2; noise(2,:)=abs(fft(A2,700)).^2; noise(3,:)=abs(fft(A3,700)).^2; noise(4,:)=abs(fft(A4,700)).^2;

%average noise PSD

noise_PSD_estimate = mean(noise,1);

%signal PSD

signal_PSD=abs(fft(model,700)).^2;

I hope I correctly calculated the PSD's. but now I am stuck. We were given that the transfer function of the wiener filter was W(f)= Sd(f)/(Sd(f)+Sn(f)) with Sd=power spectral density of the desired signal, and Sd=power spectral density of the noise

thus I created a filter b=signal_PSD; a=signal_PSD+noise_PSD;

y=filter(b,a,example_beat);

but when plotting this, It didn't give any good results....

Help is appreciated...