Scribd Logo
Upload

Welcome to Scribd!

  • 55 views

    1) Computing of The Maximum and Minimum Values of The Clean and Noisy ECG Signals

    Uploaded by

    CAKPO
    The document discusses processing an ECG signal to remove noise. It computes the minimum and maximum values of clean and noisy ECG signals. It then calculates the DC components of each signal using wavelet decomposition. It performs an FFT on the noisy signal to analyze its frequency spectrum. Finally, it designs and applies a filter to remove unwanted noise based on the frequency ranges identified in the FFT.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd

    1) Computing of The Maximum and Minimum Values of The Clean and Noisy ECG Signals

    Uploaded by

    CAKPO
    55 views6 pages
    The document discusses processing an ECG signal to remove noise. It computes the minimum and maximum values of clean and noisy ECG signals. It then calculates the DC components of each signal using wavelet decomposition. It performs an FFT on the noisy signal to analyze its frequency spectrum. Finally, it designs and applies a filter to remove unwanted noise based on the frequency ranges identified in the FFT.

    Original Title

    ecgfi

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    The document discusses processing an ECG signal to remove noise. It computes the minimum and maximum values of clean and noisy ECG signals. It then calculates the DC components of each signal using wavelet decomposition. It performs an FFT on the noisy signal to analyze its frequency spectrum. Finally, it designs and applies a filter to remove unwanted noise based on the frequency ranges identified in the FFT.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    55 views6 pages

    1) Computing of The Maximum and Minimum Values of The Clean and Noisy ECG Signals

    Uploaded by

    CAKPO
    The document discusses processing an ECG signal to remove noise. It computes the minimum and maximum values of clean and noisy ECG signals. It then calculates the DC components of each signal using wavelet decomposition. It performs an FFT on the noisy signal to analyze its frequency spectrum. Finally, it designs and applies a filter to remove unwanted noise based on the frequency ranges identified in the FFT.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    You are on page 1of 6

    1) Computing of the maximum and minimum values of the clean and noisy ECG signals

    To calculate those values: I have read before the clean ecg signal then noisy signal and we

    get their maximum and minimum. I take the ecg_004 signal as noisy singnal.

    clear

    %Fs = 200; %Sampling frequency


    [x1, Fs1] = audioread('ecgclean.wav');
    [x2, Fs2] = audioread('ecg_004.wav');

    max_ecgclean = max(x1);
    min_ecgclean = min(x1);

    fprintf('max_ecgclean = %d\n',max_ecgclean);

    max_ecgclean = 5.947876e-01

    fprintf('min_ecgclean = %d\n',min_ecgclean);

    min_ecgclean = 2.940063e-01

    max_ecg_004 = max(x2);
    min_ecg_004 = min(x2);

    fprintf('max_ecg_004 = %d\n',max_ecg_004);

    max_ecg_004 = 6.325989e-01

    fprintf('min_ecg_004 = %d\n',min_ecg_004);

    min_ecg_004 = -1.878662e-01

    xc = x1(1:1000);
    xn = x2(1:1000);

    subplot(4,1,1)
    plot(xn), title('noisy ecg signal befor DC components'), grid on
    subplot(4,1,2)
    plot(xc), title('clean ecg signal befor DC components'), grid on

    2) Computing of DC components of clean and noisy signals


    -Clean ecg

    %Question 2

    [C, L] = wavedec(xc,9,'bior3.7');
    a19 = wrcoef('a',C,L,'bior3.7',9);
    d19 = wrcoef('d',C,L,'bior3.7',9);
    d18 = wrcoef('d',C,L,'bior3.7',8);
    d17 = wrcoef('d',C,L,'bior3.7',7);
    d16 = wrcoef('d',C,L,'bior3.7',6);

    1
    d15 = wrcoef('d',C,L,'bior3.7',5);
    d14 = wrcoef('d',C,L,'bior3.7',4);
    d13 = wrcoef('d',C,L,'bior3.7',3);
    d12 = wrcoef('d',C,L,'bior3.7',2);
    d11 = wrcoef('d',C,L,'bior3.7',1);

    subplot(4,1,4)
    y11 = d19+d18+d17+d16+d15+d14+d13+d12+d11;
    plot(y11)
    title('clean ecg signal after DC components'), grid on

    -Noisy ecg

    [C, L] = wavedec(xn,9,'bior3.7');
    a29 = wrcoef('a',C,L,'bior3.7',9);
    d29 = wrcoef('d',C,L,'bior3.7',9);
    d28 = wrcoef('d',C,L,'bior3.7',8);
    d27 = wrcoef('d',C,L,'bior3.7',7);
    d26 = wrcoef('d',C,L,'bior3.7',6);
    d25 = wrcoef('d',C,L,'bior3.7',5);
    d24 = wrcoef('d',C,L,'bior3.7',4);
    d23 = wrcoef('d',C,L,'bior3.7',3);
    d22 = wrcoef('d',C,L,'bior3.7',2);
    d21 = wrcoef('d',C,L,'bior3.7',1);

    subplot(4,1,3)
    y22 = d29+d28+d27+d26+d25+d24+d23+d22+d21;
    plot(y22)
    title('noisy ecg signal after DC components'), grid on

    2
    3) Compute a 1024 point FFT from the initial 1024 samples segment of the noisy ECG
    without DC component.
    To calculate the fast fourier transform (FFT) of the noisy signal with the first 1024 points of x2

    I caclulate the frequency of x-axis betwen - and all times where N is the size

    of x2 and I use matlab function fft() to normalize x2 and then I plot.

    figure(2)
    %Question 3
    x2 = x2(1:1024);
    N = size(x2, 1);

    dx2 = Fs2 / N;
    w = (-(N/2):(N/2)-1)*dx2;
    y2 = fft(x2) / N; % For normalizing
    y3 = fftshift(y2);

    plot(w, abs(y3));
    title('Amplitude Spectrum');
    xlabel('Frequency(Hz)');

    3
    4) Remorving of unwanted noise based on the frequency spectrum obtained in (3

    I use the matlab function butter() and his arguments where :

    n: is the filter length

    beginFreq: is the noprmalized cuttof frequency

    b: the numerator of transfer function

    a: the denominator of transfer function

    %Question 4

    filtering f1
    n = 2;
    beginFreq = 45 / (Fs2/2);
    [b,a] = butter(n, beginFreq,'s');
    fOut = filter(b, a, x2);

    % extract f1
    [b,a] = butter(n, beginFreq,'low');

    4
    f1 = filter(b, a, x2);

    filtering f2
    beginFreq = 80 / (Fs2/2);
    [b,a] = butter(n, beginFreq, 's');
    fOut = filter(b, a, fOut);

    % extract f2
    [b,a] = butter(n, beginFreq, 'high');
    f2 = filter(b, a, x2);

    % For normalizing
    Ym = max(max(max(fOut)),max(abs(min(fOut))));
    fOut = fOut ./ Ym;

    After processing

    plot(w,abs(fftshift(fft(fOut)/N)));
    title('After processing: Amplitude Spectrum'); xlabel('Frequency(Hz)');

    Here to plot the signal in time domaine, I firstly calulate the period given by;

    where Fs2 is the frequency of the noisy signal then I compute the times values

    5
    which are the multiples of the period T in second between 0 and L-1 with L the length of the signal.

    figure;
    x2 = x2(1:1000);
    T = 1 / Fs2; % Sampling period
    L = length(x2); % Length of signal
    t = (0:L-1) * T; % Time vector
    N = size(x2, 1);
    stem(t, fOut(1:1000));
    title('After processing: Time-domain'); xlabel('time(seconds)');

    You might also like