|
a |
|
b/pan_tompkin.m |
|
|
1 |
function [qrs_amp_raw,qrs_i_raw,delay]=pan_tompkin(ecg,fs,gr) |
|
|
2 |
|
|
|
3 |
%% function [qrs_amp_raw,qrs_i_raw,delay]=pan_tompkin(ecg,fs) |
|
|
4 |
% Complete implementation of Pan-Tompkins algorithm |
|
|
5 |
|
|
|
6 |
%% Inputs |
|
|
7 |
% ecg : raw ecg vector signal 1d signal |
|
|
8 |
% fs : sampling frequency e.g. 200Hz, 400Hz and etc |
|
|
9 |
% gr : flag to plot or not plot (set it 1 to have a plot or set it zero not |
|
|
10 |
% to see any plots |
|
|
11 |
%% Outputs |
|
|
12 |
% qrs_amp_raw : amplitude of R waves amplitudes |
|
|
13 |
% qrs_i_raw : index of R waves |
|
|
14 |
% delay : number of samples which the signal is delayed due to the |
|
|
15 |
% filtering |
|
|
16 |
%% Method : |
|
|
17 |
|
|
|
18 |
%% PreProcessing |
|
|
19 |
% 1) Signal is preprocessed , if the sampling frequency is higher then it is downsampled |
|
|
20 |
% and if it is lower upsampled to make the sampling frequency 200 Hz |
|
|
21 |
% with the same filtering setups introduced in Pan |
|
|
22 |
% tompkins paper (a combination of low pass and high pass filter 5-15 Hz) |
|
|
23 |
% to get rid of the baseline wander and muscle noise. |
|
|
24 |
|
|
|
25 |
% 2) The filtered signal |
|
|
26 |
% is derivated using a derivating filter to high light the QRS complex. |
|
|
27 |
|
|
|
28 |
% 3) Signal is squared.4)Signal is averaged with a moving window to get rid |
|
|
29 |
% of noise (0.150 seconds length). |
|
|
30 |
|
|
|
31 |
% 5) depending on the sampling frequency of your signal the filtering |
|
|
32 |
% options are changed to best match the characteristics of your ecg signal |
|
|
33 |
|
|
|
34 |
% 6) Unlike the other implementations in this implementation the desicion |
|
|
35 |
% rule of the Pan tompkins is implemented completely. |
|
|
36 |
|
|
|
37 |
%% Decision Rule |
|
|
38 |
% At this point in the algorithm, the preceding stages have produced a roughly pulse-shaped |
|
|
39 |
% waveform at the output of the MWI . The determination as to whether this pulse |
|
|
40 |
% corresponds to a QRS complex (as opposed to a high-sloped T-wave or a noise artefact) is |
|
|
41 |
% performed with an adaptive thresholding operation and other decision |
|
|
42 |
% rules outlined below; |
|
|
43 |
|
|
|
44 |
% a) FIDUCIAL MARK - The waveform is first processed to produce a set of weighted unit |
|
|
45 |
% samples at the location of the MWI maxima. This is done in order to localize the QRS |
|
|
46 |
% complex to a single instant of time. The w[k] weighting is the maxima value. |
|
|
47 |
|
|
|
48 |
% b) THRESHOLDING - When analyzing the amplitude of the MWI output, the algorithm uses |
|
|
49 |
% two threshold values (THR_SIG and THR_NOISE, appropriately initialized during a brief |
|
|
50 |
% 2 second training phase) that continuously adapt to changing ECG signal quality. The |
|
|
51 |
% first pass through y[n] uses these thresholds to classify the each non-zero sample |
|
|
52 |
% (CURRENTPEAK) as either signal or noise: |
|
|
53 |
% If CURRENTPEAK > THR_SIG, that location is identified as a QRS complex |
|
|
54 |
% candidate?and the signal level (SIG_LEV) is updated: |
|
|
55 |
% SIG _ LEV = 0.125 CURRENTPEAK + 0.875?SIG _ LEV |
|
|
56 |
|
|
|
57 |
% If THR_NOISE < CURRENTPEAK < THR_SIG, then that location is identified as a |
|
|
58 |
% Noise peak?and the noise level (NOISE_LEV) is updated: |
|
|
59 |
% NOISE _ LEV = 0.125CURRENTPEAK + 0.875?NOISE _ LEV |
|
|
60 |
% Based on new estimates of the signal and noise levels (SIG_LEV and NOISE_LEV, |
|
|
61 |
% respectively) at that point in the ECG, the thresholds are adjusted as follows: |
|
|
62 |
% THR _ SIG = NOISE _ LEV + 0.25 ?(SIG _ LEV-NOISE _ LEV ) |
|
|
63 |
% THR _ NOISE = 0.5?(THR _ SIG) |
|
|
64 |
% These adjustments lower the threshold gradually in signal segments that are deemed to |
|
|
65 |
% be of poorer quality. |
|
|
66 |
|
|
|
67 |
|
|
|
68 |
% c) SEARCHBACK FOR MISSED QRS COMPLEXES - In the thresholding step above, if |
|
|
69 |
% CURRENTPEAK < THR_SIG, the peak is deemed not to have resulted from a QRS |
|
|
70 |
% complex. If however, an unreasonably long period has expired without an abovethreshold |
|
|
71 |
% peak, the algorithm will assume a QRS has been missed and perform a |
|
|
72 |
% searchback. This limits the number of false negatives. The minimum time used to trigger |
|
|
73 |
% a searchback is 1.66 times the current R peak to R peak time period (called the RR |
|
|
74 |
% interval). This value has a physiological origin - the time value between adjacent |
|
|
75 |
% heartbeats cannot change more quickly than this. The missed QRS complex is assumed |
|
|
76 |
% to occur at the location of the highest peak in the interval that lies between THR_SIG and |
|
|
77 |
% THR_NOISE. In this algorithm, two average RR intervals are stored,the first RR interval is |
|
|
78 |
% calculated as an average of the last eight QRS locations in order to adapt to changing heart |
|
|
79 |
% rate and the second RR interval mean is the mean |
|
|
80 |
% of the most regular RR intervals . The threshold is lowered if the heart rate is not regular |
|
|
81 |
% to improve detection. |
|
|
82 |
|
|
|
83 |
% d) ELIMINATION OF MULTIPLE DETECTIONS WITHIN REFRACTORY PERIOD - It is |
|
|
84 |
% impossible for a legitimate QRS complex to occur if it lies within 200ms after a previously |
|
|
85 |
% detected one. This constraint is a physiological one ?due to the refractory period during |
|
|
86 |
% which ventricular depolarization cannot occur despite a stimulus[1]. As QRS complex |
|
|
87 |
% candidates are generated, the algorithm eliminates such physically impossible events, |
|
|
88 |
% thereby reducing false positives. |
|
|
89 |
|
|
|
90 |
% e) T WAVE DISCRIMINATION - Finally, if a QRS candidate occurs after the 200ms |
|
|
91 |
% refractory period but within 360ms of the previous QRS, the algorithm determines |
|
|
92 |
% whether this is a genuine QRS complex of the next heartbeat or an abnormally prominent |
|
|
93 |
% T wave. This decision is based on the mean slope of the waveform at that position. A slope of |
|
|
94 |
% less than one half that of the previous QRS complex is consistent with the slower |
|
|
95 |
% changing behaviour of a T wave ?otherwise, it becomes a QRS detection. |
|
|
96 |
% Extra concept : beside the points mentioned in the paper, this code also |
|
|
97 |
% checks if the occured peak which is less than 360 msec latency has also a |
|
|
98 |
% latency less than 0,5*mean_RR if yes this is counted as noise |
|
|
99 |
|
|
|
100 |
% f) In the final stage , the output of R waves detected in smoothed signal is analyzed and double |
|
|
101 |
% checked with the help of the output of the bandpass signal to improve |
|
|
102 |
% detection and find the original index of the real R waves on the raw ecg |
|
|
103 |
% signal |
|
|
104 |
|
|
|
105 |
%% References : |
|
|
106 |
|
|
|
107 |
%[1]PAN.J, TOMPKINS. W.J,"A Real-Time QRS Detection Algorithm" IEEE |
|
|
108 |
%TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. BME-32, NO. 3, MARCH 1985. |
|
|
109 |
|
|
|
110 |
%% Author : Hooman Sedghamiz |
|
|
111 |
% Linkoping university |
|
|
112 |
% email : hoose792@student.liu.se |
|
|
113 |
% hooman.sedghamiz@medel.com |
|
|
114 |
|
|
|
115 |
% Any direct or indirect use of this code should be referenced |
|
|
116 |
% Copyright march 2014 |
|
|
117 |
%% |
|
|
118 |
if ~isvector(ecg) |
|
|
119 |
error('ecg must be a row or column vector'); |
|
|
120 |
end |
|
|
121 |
|
|
|
122 |
|
|
|
123 |
if nargin < 3 |
|
|
124 |
gr = 1; % on default the function always plots |
|
|
125 |
end |
|
|
126 |
ecg = ecg(:); % vectorize |
|
|
127 |
|
|
|
128 |
%% Initialize |
|
|
129 |
qrs_c =[]; %amplitude of R |
|
|
130 |
qrs_i =[]; %index |
|
|
131 |
SIG_LEV = 0; |
|
|
132 |
nois_c =[]; |
|
|
133 |
nois_i =[]; |
|
|
134 |
delay = 0; |
|
|
135 |
skip = 0; % becomes one when a T wave is detected |
|
|
136 |
not_nois = 0; % it is not noise when not_nois = 1 |
|
|
137 |
selected_RR =[]; % Selected RR intervals |
|
|
138 |
m_selected_RR = 0; |
|
|
139 |
mean_RR = 0; |
|
|
140 |
qrs_i_raw =[]; |
|
|
141 |
qrs_amp_raw=[]; |
|
|
142 |
ser_back = 0; |
|
|
143 |
test_m = 0; |
|
|
144 |
SIGL_buf = []; |
|
|
145 |
NOISL_buf = []; |
|
|
146 |
THRS_buf = []; |
|
|
147 |
SIGL_buf1 = []; |
|
|
148 |
NOISL_buf1 = []; |
|
|
149 |
THRS_buf1 = []; |
|
|
150 |
|
|
|
151 |
|
|
|
152 |
%% Plot differently based on filtering settings |
|
|
153 |
if gr |
|
|
154 |
if fs == 200 |
|
|
155 |
figure, ax(1)=subplot(321);plot(ecg);axis tight;title('Raw ECG Signal'); |
|
|
156 |
else |
|
|
157 |
figure, ax(1)=subplot(3,2,[1 2]);plot(ecg);axis tight;title('Raw ECG Signal'); |
|
|
158 |
end |
|
|
159 |
end |
|
|
160 |
%% Noise cancelation(Filtering) % Filters (Filter in between 5-15 Hz) |
|
|
161 |
if fs == 200 |
|
|
162 |
%% Low Pass Filter H(z) = ((1 - z^(-6))^2)/(1 - z^(-1))^2 |
|
|
163 |
b = [1 0 0 0 0 0 -2 0 0 0 0 0 1]; |
|
|
164 |
a = [1 -2 1]; |
|
|
165 |
h_l = filter(b,a,[1 zeros(1,12)]); |
|
|
166 |
ecg_l = conv (ecg ,h_l); |
|
|
167 |
ecg_l = ecg_l/ max( abs(ecg_l)); |
|
|
168 |
delay = 6; %based on the paper |
|
|
169 |
if gr |
|
|
170 |
ax(2)=subplot(322);plot(ecg_l);axis tight;title('Low pass filtered'); |
|
|
171 |
end |
|
|
172 |
%% High Pass filter H(z) = (-1+32z^(-16)+z^(-32))/(1+z^(-1)) |
|
|
173 |
b = [-1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 32 -32 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1]; |
|
|
174 |
a = [1 -1]; |
|
|
175 |
h_h = filter(b,a,[1 zeros(1,32)]); |
|
|
176 |
ecg_h = conv (ecg_l ,h_h); |
|
|
177 |
ecg_h = ecg_h/ max( abs(ecg_h)); |
|
|
178 |
delay = delay + 16; % 16 samples for highpass filtering |
|
|
179 |
if gr |
|
|
180 |
ax(3)=subplot(323);plot(ecg_h);axis tight;title('High Pass Filtered'); |
|
|
181 |
end |
|
|
182 |
else |
|
|
183 |
%% bandpass filter for Noise cancelation of other sampling frequencies(Filtering) |
|
|
184 |
f1=5; %cuttoff low frequency to get rid of baseline wander |
|
|
185 |
f2=15; %cuttoff frequency to discard high frequency noise |
|
|
186 |
Wn=[f1 f2]*2/fs; % cutt off based on fs |
|
|
187 |
N = 3; % order of 3 less processing |
|
|
188 |
[a,b] = butter(N,Wn); %bandpass filtering |
|
|
189 |
ecg_h = filtfilt(a,b,ecg); |
|
|
190 |
ecg_h = ecg_h/ max( abs(ecg_h)); |
|
|
191 |
if gr |
|
|
192 |
ax(3)=subplot(323);plot(ecg_h);axis tight;title('Band Pass Filtered'); |
|
|
193 |
end |
|
|
194 |
end |
|
|
195 |
%% derivative filter H(z) = (1/8T)(-z^(-2) - 2z^(-1) + 2z + z^(2)) |
|
|
196 |
h_d = [-1 -2 0 2 1]*(1/8);%1/8*fs |
|
|
197 |
ecg_d = conv (ecg_h ,h_d); |
|
|
198 |
ecg_d = ecg_d/max(ecg_d); |
|
|
199 |
delay = delay + 2; % delay of derivative filter 2 samples |
|
|
200 |
if gr |
|
|
201 |
ax(4)=subplot(324);plot(ecg_d);axis tight;title('Filtered with the derivative filter'); |
|
|
202 |
end |
|
|
203 |
%% Squaring nonlinearly enhance the dominant peaks |
|
|
204 |
ecg_s = ecg_d.^2; |
|
|
205 |
if gr |
|
|
206 |
ax(5)=subplot(325);plot(ecg_s);axis tight;title('Squared'); |
|
|
207 |
end |
|
|
208 |
|
|
|
209 |
|
|
|
210 |
|
|
|
211 |
%% Moving average Y(nt) = (1/N)[x(nT-(N - 1)T)+ x(nT - (N - 2)T)+...+x(nT)] |
|
|
212 |
ecg_m = conv(ecg_s ,ones(1 ,round(0.150*fs))/round(0.150*fs)); |
|
|
213 |
delay = delay + 15; |
|
|
214 |
|
|
|
215 |
if gr |
|
|
216 |
ax(6)=subplot(326);plot(ecg_m);axis tight;title('Averaged with 30 samples length,Black noise,Green Adaptive Threshold,RED Sig Level,Red circles QRS adaptive threshold'); |
|
|
217 |
axis tight; |
|
|
218 |
end |
|
|
219 |
|
|
|
220 |
%% Fiducial Mark |
|
|
221 |
% Note : a minimum distance of 40 samples is considered between each R wave |
|
|
222 |
% since in physiological point of view no RR wave can occur in less than |
|
|
223 |
% 200 msec distance |
|
|
224 |
[pks,locs] = findpeaks(ecg_m,'MINPEAKDISTANCE',round(0.2*fs)); |
|
|
225 |
|
|
|
226 |
|
|
|
227 |
|
|
|
228 |
|
|
|
229 |
%% initialize the training phase (2 seconds of the signal) to determine the THR_SIG and THR_NOISE |
|
|
230 |
THR_SIG = max(ecg_m(1:2*fs))*1/3; % 0.25 of the max amplitude |
|
|
231 |
THR_NOISE = mean(ecg_m(1:2*fs))*1/2; % 0.5 of the mean signal is considered to be noise |
|
|
232 |
SIG_LEV= THR_SIG; |
|
|
233 |
NOISE_LEV = THR_NOISE; |
|
|
234 |
|
|
|
235 |
|
|
|
236 |
%% Initialize bandpath filter threshold(2 seconds of the bandpass signal) |
|
|
237 |
THR_SIG1 = max(ecg_h(1:2*fs))*1/3; % 0.25 of the max amplitude |
|
|
238 |
THR_NOISE1 = mean(ecg_h(1:2*fs))*1/2; % |
|
|
239 |
SIG_LEV1 = THR_SIG1; % Signal level in Bandpassed filter |
|
|
240 |
NOISE_LEV1 = THR_NOISE1; % Noise level in Bandpassed filter |
|
|
241 |
%% Thresholding and online desicion rule |
|
|
242 |
|
|
|
243 |
for i = 1 : length(pks) |
|
|
244 |
|
|
|
245 |
%% locate the corresponding peak in the filtered signal |
|
|
246 |
if locs(i)-round(0.150*fs)>= 1 && locs(i)<= length(ecg_h) |
|
|
247 |
[y_i x_i] = max(ecg_h(locs(i)-round(0.150*fs):locs(i))); |
|
|
248 |
else |
|
|
249 |
if i == 1 |
|
|
250 |
[y_i x_i] = max(ecg_h(1:locs(i))); |
|
|
251 |
ser_back = 1; |
|
|
252 |
elseif locs(i)>= length(ecg_h) |
|
|
253 |
[y_i x_i] = max(ecg_h(locs(i)-round(0.150*fs):end)); |
|
|
254 |
end |
|
|
255 |
|
|
|
256 |
end |
|
|
257 |
|
|
|
258 |
|
|
|
259 |
%% update the heart_rate (Two heart rate means one the moste recent and the other selected) |
|
|
260 |
if length(qrs_c) >= 9 |
|
|
261 |
|
|
|
262 |
diffRR = diff(qrs_i(end-8:end)); %calculate RR interval |
|
|
263 |
mean_RR = mean(diffRR); % calculate the mean of 8 previous R waves interval |
|
|
264 |
comp =qrs_i(end)-qrs_i(end-1); %latest RR |
|
|
265 |
if comp <= 0.92*mean_RR || comp >= 1.16*mean_RR |
|
|
266 |
% lower down thresholds to detect better in MVI |
|
|
267 |
THR_SIG = 0.5*(THR_SIG); |
|
|
268 |
%THR_NOISE = 0.5*(THR_SIG); |
|
|
269 |
% lower down thresholds to detect better in Bandpass filtered |
|
|
270 |
THR_SIG1 = 0.5*(THR_SIG1); |
|
|
271 |
%THR_NOISE1 = 0.5*(THR_SIG1); |
|
|
272 |
|
|
|
273 |
else |
|
|
274 |
m_selected_RR = mean_RR; %the latest regular beats mean |
|
|
275 |
end |
|
|
276 |
|
|
|
277 |
end |
|
|
278 |
|
|
|
279 |
%% calculate the mean of the last 8 R waves to make sure that QRS is not |
|
|
280 |
% missing(If no R detected , trigger a search back) 1.66*mean |
|
|
281 |
|
|
|
282 |
if m_selected_RR |
|
|
283 |
test_m = m_selected_RR; %if the regular RR availabe use it |
|
|
284 |
elseif mean_RR && m_selected_RR == 0 |
|
|
285 |
test_m = mean_RR; |
|
|
286 |
else |
|
|
287 |
test_m = 0; |
|
|
288 |
end |
|
|
289 |
|
|
|
290 |
if test_m |
|
|
291 |
if (locs(i) - qrs_i(end)) >= round(1.66*test_m)% it shows a QRS is missed |
|
|
292 |
[pks_temp,locs_temp] = max(ecg_m(qrs_i(end)+ round(0.200*fs):locs(i)-round(0.200*fs))); % search back and locate the max in this interval |
|
|
293 |
locs_temp = qrs_i(end)+ round(0.200*fs) + locs_temp -1; %location |
|
|
294 |
|
|
|
295 |
if pks_temp > THR_NOISE |
|
|
296 |
qrs_c = [qrs_c pks_temp]; |
|
|
297 |
qrs_i = [qrs_i locs_temp]; |
|
|
298 |
|
|
|
299 |
% find the location in filtered sig |
|
|
300 |
if locs_temp <= length(ecg_h) |
|
|
301 |
[y_i_t x_i_t] = max(ecg_h(locs_temp-round(0.150*fs):locs_temp)); |
|
|
302 |
else |
|
|
303 |
[y_i_t x_i_t] = max(ecg_h(locs_temp-round(0.150*fs):end)); |
|
|
304 |
end |
|
|
305 |
% take care of bandpass signal threshold |
|
|
306 |
if y_i_t > THR_NOISE1 |
|
|
307 |
|
|
|
308 |
qrs_i_raw = [qrs_i_raw locs_temp-round(0.150*fs)+ (x_i_t - 1)];% save index of bandpass |
|
|
309 |
qrs_amp_raw =[qrs_amp_raw y_i_t]; %save amplitude of bandpass |
|
|
310 |
SIG_LEV1 = 0.25*y_i_t + 0.75*SIG_LEV1; %when found with the second thres |
|
|
311 |
end |
|
|
312 |
|
|
|
313 |
not_nois = 1; |
|
|
314 |
SIG_LEV = 0.25*pks_temp + 0.75*SIG_LEV ; %when found with the second threshold |
|
|
315 |
end |
|
|
316 |
|
|
|
317 |
else |
|
|
318 |
not_nois = 0; |
|
|
319 |
|
|
|
320 |
end |
|
|
321 |
end |
|
|
322 |
|
|
|
323 |
|
|
|
324 |
|
|
|
325 |
|
|
|
326 |
%% find noise and QRS peaks |
|
|
327 |
if pks(i) >= THR_SIG |
|
|
328 |
|
|
|
329 |
% if a QRS candidate occurs within 360ms of the previous QRS |
|
|
330 |
% ,the algorithm determines if its T wave or QRS |
|
|
331 |
if length(qrs_c) >= 3 |
|
|
332 |
if (locs(i)-qrs_i(end)) <= round(0.3600*fs) |
|
|
333 |
Slope1 = mean(diff(ecg_m(locs(i)-round(0.075*fs):locs(i)))); %mean slope of the waveform at that position |
|
|
334 |
Slope2 = mean(diff(ecg_m(qrs_i(end)-round(0.075*fs):qrs_i(end)))); %mean slope of previous R wave |
|
|
335 |
if abs(Slope1) <= abs(0.5*(Slope2)) % slope less then 0.5 of previous R |
|
|
336 |
nois_c = [nois_c pks(i)]; |
|
|
337 |
nois_i = [nois_i locs(i)]; |
|
|
338 |
skip = 1; % T wave identification |
|
|
339 |
% adjust noise level in both filtered and |
|
|
340 |
% MVI |
|
|
341 |
NOISE_LEV1 = 0.125*y_i + 0.875*NOISE_LEV1; |
|
|
342 |
NOISE_LEV = 0.125*pks(i) + 0.875*NOISE_LEV; |
|
|
343 |
else |
|
|
344 |
skip = 0; |
|
|
345 |
end |
|
|
346 |
|
|
|
347 |
end |
|
|
348 |
end |
|
|
349 |
|
|
|
350 |
if skip == 0 % skip is 1 when a T wave is detected |
|
|
351 |
qrs_c = [qrs_c pks(i)]; |
|
|
352 |
qrs_i = [qrs_i locs(i)]; |
|
|
353 |
|
|
|
354 |
% bandpass filter check threshold |
|
|
355 |
if y_i >= THR_SIG1 |
|
|
356 |
if ser_back |
|
|
357 |
qrs_i_raw = [qrs_i_raw x_i]; % save index of bandpass |
|
|
358 |
else |
|
|
359 |
qrs_i_raw = [qrs_i_raw locs(i)-round(0.150*fs)+ (x_i - 1)];% save index of bandpass |
|
|
360 |
end |
|
|
361 |
qrs_amp_raw =[qrs_amp_raw y_i];% save amplitude of bandpass |
|
|
362 |
SIG_LEV1 = 0.125*y_i + 0.875*SIG_LEV1;% adjust threshold for bandpass filtered sig |
|
|
363 |
end |
|
|
364 |
|
|
|
365 |
% adjust Signal level |
|
|
366 |
SIG_LEV = 0.125*pks(i) + 0.875*SIG_LEV ; |
|
|
367 |
end |
|
|
368 |
|
|
|
369 |
|
|
|
370 |
elseif THR_NOISE <= pks(i) && pks(i)<THR_SIG |
|
|
371 |
|
|
|
372 |
%adjust Noise level in filtered sig |
|
|
373 |
NOISE_LEV1 = 0.125*y_i + 0.875*NOISE_LEV1; |
|
|
374 |
%adjust Noise level in MVI |
|
|
375 |
NOISE_LEV = 0.125*pks(i) + 0.875*NOISE_LEV; |
|
|
376 |
|
|
|
377 |
|
|
|
378 |
|
|
|
379 |
elseif pks(i) < THR_NOISE |
|
|
380 |
nois_c = [nois_c pks(i)]; |
|
|
381 |
nois_i = [nois_i locs(i)]; |
|
|
382 |
|
|
|
383 |
% noise level in filtered signal |
|
|
384 |
NOISE_LEV1 = 0.125*y_i + 0.875*NOISE_LEV1; |
|
|
385 |
%end |
|
|
386 |
|
|
|
387 |
%adjust Noise level in MVI |
|
|
388 |
NOISE_LEV = 0.125*pks(i) + 0.875*NOISE_LEV; |
|
|
389 |
|
|
|
390 |
|
|
|
391 |
end |
|
|
392 |
|
|
|
393 |
|
|
|
394 |
|
|
|
395 |
|
|
|
396 |
|
|
|
397 |
%% adjust the threshold with SNR |
|
|
398 |
if NOISE_LEV ~= 0 || SIG_LEV ~= 0 |
|
|
399 |
THR_SIG = NOISE_LEV + 0.25*(abs(SIG_LEV - NOISE_LEV)); |
|
|
400 |
THR_NOISE = 0.5*(THR_SIG); |
|
|
401 |
end |
|
|
402 |
|
|
|
403 |
% adjust the threshold with SNR for bandpassed signal |
|
|
404 |
if NOISE_LEV1 ~= 0 || SIG_LEV1 ~= 0 |
|
|
405 |
THR_SIG1 = NOISE_LEV1 + 0.25*(abs(SIG_LEV1 - NOISE_LEV1)); |
|
|
406 |
THR_NOISE1 = 0.5*(THR_SIG1); |
|
|
407 |
end |
|
|
408 |
|
|
|
409 |
|
|
|
410 |
% take a track of thresholds of smoothed signal |
|
|
411 |
SIGL_buf = [SIGL_buf SIG_LEV]; |
|
|
412 |
NOISL_buf = [NOISL_buf NOISE_LEV]; |
|
|
413 |
THRS_buf = [THRS_buf THR_SIG]; |
|
|
414 |
|
|
|
415 |
% take a track of thresholds of filtered signal |
|
|
416 |
SIGL_buf1 = [SIGL_buf1 SIG_LEV1]; |
|
|
417 |
NOISL_buf1 = [NOISL_buf1 NOISE_LEV1]; |
|
|
418 |
THRS_buf1 = [THRS_buf1 THR_SIG1]; |
|
|
419 |
|
|
|
420 |
|
|
|
421 |
|
|
|
422 |
|
|
|
423 |
skip = 0; %reset parameters |
|
|
424 |
not_nois = 0; %reset parameters |
|
|
425 |
ser_back = 0; %reset bandpass param |
|
|
426 |
end |
|
|
427 |
|
|
|
428 |
if gr |
|
|
429 |
hold on,scatter(qrs_i,qrs_c,'m'); |
|
|
430 |
hold on,plot(locs,NOISL_buf,'--k','LineWidth',2); |
|
|
431 |
hold on,plot(locs,SIGL_buf,'--r','LineWidth',2); |
|
|
432 |
hold on,plot(locs,THRS_buf,'--g','LineWidth',2); |
|
|
433 |
if ax(:) |
|
|
434 |
linkaxes(ax,'x'); |
|
|
435 |
zoom on; |
|
|
436 |
end |
|
|
437 |
end |
|
|
438 |
|
|
|
439 |
|
|
|
440 |
|
|
|
441 |
|
|
|
442 |
%% overlay on the signals |
|
|
443 |
if gr |
|
|
444 |
figure,az(1)=subplot(311);plot(ecg_h);title('QRS on Filtered Signal');axis tight; |
|
|
445 |
hold on,scatter(qrs_i_raw,qrs_amp_raw,'m'); |
|
|
446 |
hold on,plot(locs,NOISL_buf1,'LineWidth',2,'Linestyle','--','color','k'); |
|
|
447 |
hold on,plot(locs,SIGL_buf1,'LineWidth',2,'Linestyle','-.','color','r'); |
|
|
448 |
hold on,plot(locs,THRS_buf1,'LineWidth',2,'Linestyle','-.','color','g'); |
|
|
449 |
az(2)=subplot(312);plot(ecg_m);title('QRS on MVI signal and Noise level(black),Signal Level (red) and Adaptive Threshold(green)');axis tight; |
|
|
450 |
hold on,scatter(qrs_i,qrs_c,'m'); |
|
|
451 |
hold on,plot(locs,NOISL_buf,'LineWidth',2,'Linestyle','--','color','k'); |
|
|
452 |
hold on,plot(locs,SIGL_buf,'LineWidth',2,'Linestyle','-.','color','r'); |
|
|
453 |
hold on,plot(locs,THRS_buf,'LineWidth',2,'Linestyle','-.','color','g'); |
|
|
454 |
az(3)=subplot(313);plot(ecg-mean(ecg));title('Pulse train of the found QRS on ECG signal');axis tight; |
|
|
455 |
line(repmat(qrs_i_raw,[2 1]),repmat([min(ecg-mean(ecg))/2; max(ecg-mean(ecg))/2],size(qrs_i_raw)),'LineWidth',2.5,'LineStyle','-.','Color','r'); |
|
|
456 |
linkaxes(az,'x'); |
|
|
457 |
zoom on; |
|
|
458 |
end |
|
|
459 |
end |
|
|
460 |
|
|
|
461 |
|
|
|
462 |
|
|
|
463 |
|
|
|
464 |
|
|
|
465 |
|
|
|
466 |
|
|
|
467 |
|
|
|
468 |
|
|
|
469 |
|