CN110477906B - Electrocardiosignal QRS wave start and stop point positioning method - Google Patents
Electrocardiosignal QRS wave start and stop point positioning method Download PDFInfo
- Publication number
- CN110477906B CN110477906B CN201910793288.5A CN201910793288A CN110477906B CN 110477906 B CN110477906 B CN 110477906B CN 201910793288 A CN201910793288 A CN 201910793288A CN 110477906 B CN110477906 B CN 110477906B
- Authority
- CN
- China
- Prior art keywords
- point
- candidate
- points
- qrs wave
- amplitude
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 60
- 238000000718 qrs complex Methods 0.000 claims abstract description 39
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 14
- 238000005516 engineering process Methods 0.000 claims abstract description 8
- 238000005070 sampling Methods 0.000 claims description 37
- 238000012216 screening Methods 0.000 claims description 18
- 230000000694 effects Effects 0.000 claims description 9
- 239000012634 fragment Substances 0.000 claims description 9
- 230000008569 process Effects 0.000 claims description 8
- 238000012360 testing method Methods 0.000 claims description 8
- 238000001914 filtration Methods 0.000 claims description 7
- 101100129590 Schizosaccharomyces pombe (strain 972 / ATCC 24843) mcp5 gene Proteins 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000007667 floating Methods 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 3
- NAWXUBYGYWOOIX-SFHVURJKSA-N (2s)-2-[[4-[2-(2,4-diaminoquinazolin-6-yl)ethyl]benzoyl]amino]-4-methylidenepentanedioic acid Chemical compound C1=CC2=NC(N)=NC(N)=C2C=C1CCC1=CC=C(C(=O)N[C@@H](CC(=C)C(O)=O)C(O)=O)C=C1 NAWXUBYGYWOOIX-SFHVURJKSA-N 0.000 claims description 2
- 230000000747 cardiac effect Effects 0.000 claims description 2
- 238000000605 extraction Methods 0.000 claims 1
- 230000008901 benefit Effects 0.000 abstract description 8
- 230000008859 change Effects 0.000 abstract description 3
- 230000003044 adaptive effect Effects 0.000 abstract description 2
- 238000001514 detection method Methods 0.000 description 12
- 230000035945 sensitivity Effects 0.000 description 5
- 230000002159 abnormal effect Effects 0.000 description 3
- 238000011426 transformation method Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000000717 retained effect Effects 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 230000002861 ventricular Effects 0.000 description 2
- 206010006578 Bundle-Branch Block Diseases 0.000 description 1
- 208000024172 Cardiovascular disease Diseases 0.000 description 1
- 101100289792 Squirrel monkey polyomavirus large T gene Proteins 0.000 description 1
- 206010047295 Ventricular hypertrophy Diseases 0.000 description 1
- 230000005856 abnormality Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 230000005284 excitation Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000003862 health status Effects 0.000 description 1
- 208000019622 heart disease Diseases 0.000 description 1
- 230000000877 morphologic effect Effects 0.000 description 1
- 238000013102 re-test Methods 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 230000034225 regulation of ventricular cardiomyocyte membrane depolarization Effects 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/318—Heart-related electrical modalities, e.g. electrocardiography [ECG]
- A61B5/346—Analysis of electrocardiograms
- A61B5/349—Detecting specific parameters of the electrocardiograph cycle
- A61B5/366—Detecting abnormal QRS complex, e.g. widening
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Cardiology (AREA)
- Heart & Thoracic Surgery (AREA)
- Molecular Biology (AREA)
- Pathology (AREA)
- Engineering & Computer Science (AREA)
- Biomedical Technology (AREA)
- Physics & Mathematics (AREA)
- Medical Informatics (AREA)
- Biophysics (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Veterinary Medicine (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention discloses a method for positioning the start point and the stop point of a QRS wave of an electrocardiosignal, which introduces an improved R wave identification technology based on amplitude and slope, an improved polygon approximation theory based on Douglas-Peucker algorithm and an adaptive threshold value theory, solves the problem of low accuracy of positioning the start point and the stop point of a QRS wave group under the conditions of ECG form change and high noise in the prior art, and keeps more candidate start points and stop points of the QRS wave group when the QRS wave group form is more complex. Other candidate points provide information on the Q peak and S peak positions, except for the start point and the end point. Especially when multiple peaks such as R 'and S' appear in the QRS complex, the invention can also identify and locate the QRS complex. This advantage of the present invention is not available with other methods.
Description
Technical Field
The invention relates to the technical field of medical electronics, in particular to a method for positioning a start point and a stop point of a QRS wave of an electrocardiosignal.
Background
Electrocardiogram (ECG) is a weak bioelectric signal generated by the heart and collected by a special device, which reflects the changes in the generation, conduction and recovery of the excitation of the heart. Because of its ability to provide information about the electrical activity of the heart, it is now widely used for the safety assessment and identification of the heart and for the diagnosis of certain cardiovascular diseases. The ECG consists of a series of repeated heart beats. Generally, each heart beat includes characteristic waves such as a P wave, a QRS complex and a T wave, and in some cases, a U wave which is in the same direction as the T wave appears on the back surface of the T wave.
The QRS complex is the most prominent characteristic wave in ECG and represents the potential changes occurring during ventricular septal, left and right ventricular depolarization. The QRS wave width (time limit) is closely related to the ventricular activity state and can reflect the health status of the heart. Normal QRS width (time limit) is 0.06s-0.10s for adults and 0.04s-0.08s for children, beyond which limit abnormalities are detected. For example, the width of the QRS complex increases when there is bundle branch block or ventricular hypertrophy in the heart. The QRS wave width (time limit) can be calculated by positioning the QRS wave start and stop points, so that the method has important practical application significance.
The ECG is easy to be interfered by noise in the acquisition process, and particularly, the noise interference on the single-lead dynamic electrocardiogram acquired by the wearable electrocardiogram equipment is larger. Some heart diseases are also reflected in the morphology of the electrocardiogram. In addition to the individual differences between people, the morphology of the ECG varies, which makes it extremely difficult to locate the start and stop points of the QRS complex. For decades, many researchers have made a great deal of effort in the location of the onset and cessation of QRS complexes. Before the onset of the QRS complex, it is often necessary to determine the heart beat of the ECG. Accurate positioning of the R wave of the QRS complex is the most straightforward method to determine the ECG beat. On the R-wave positioning method, a differential threshold method, a time-frequency method, a morphological method, and the like have been proposed. The methods mainly utilize the form of ECG or the characteristics on the frequency domain to position R waves, can obtain satisfactory positioning effect under the condition of low noise, and the highest accuracy can reach more than 99 percent. The proposed methods for locating the start and stop points of the QRS complex mainly include envelope-based methods, frequency domain transformation methods, local transformation methods, and the like. The method based on the envelope has the advantages of high real-time performance, but has higher requirements on the regularity of the QRS complex morphology, and has low accuracy of positioning the QRS complex start and stop points when the abnormal conditions such as large T wave width exist. The frequency domain transformation method comprises filtering, empirical mode decomposition, wavelet transformation and the like. The method has the advantages of strong noise resistance, but has obvious defects, which are mainly reflected in that the calculation amount is large, the calculation is complex, and the position deviation of a boundary point is easily introduced after the electrocardiogram is transformed. The partial transform method is generally computationally simple, but is relatively sensitive to noise. In conclusion, the error of positioning the QRS complex heart beat position and the start and stop point is large when the QRS wave shape changes greatly and the noise is large in the existing method.
Disclosure of Invention
Aiming at the problems, the invention provides a method for positioning the QRS wave start and stop points of electrocardiosignals, which introduces an improved R wave identification technology based on amplitude and slope, an improved polygon approximation theory based on Douglas-Peucker algorithm and an adaptive threshold value theory so as to solve the problem of low accuracy in positioning the QRS wave group start and stop points under the conditions of ECG form change and high noise in the prior art.
The invention realizes the purpose through the following technical scheme:
a method for positioning a start point and a stop point of a QRS wave of an electrocardiosignal comprises the following steps:
step one, preprocessing an ECG signal;
the pretreatment comprises the following steps: filtering out baseline wander in the ECG by adopting a median filtering technology; to reduce the effect of floating point operations, the filtered ECG is scaled 500 times, denoted by signal (n); wherein n represents a sampling point; for an ECG signal of length N, then N is 1,2, …, N;
step two, detecting the ECG heartbeat position; piecewise fitting the signal (n) by using a least square quadratic polynomial technology, and reconstructing the signal (n) by using the slope and amplitude obtained by fitting so as to enhance the anti-noise capability of the signal; the reconstructed signal (n) is represented by s1(n), and the sampling points of the two samples are in one-to-one correspondence; extracting sampling points corresponding to signals with the amplitude larger than a threshold value from s1(n), and recording the sampling points into an R _ candidate sequence; calculating the difference of the R _ candidate sequence to obtain a difference sequence DR _ candidate; identifying from the DR _ candidate sequence that the differential value is greater than 0.25 fsThe sampling point of (1) and its immediately preceding sampling point; detecting the amplitude maximum value of s1(n) between all two sampling point pairs, and detecting the sampling point corresponding to the maximum value as the R peak position, namely the heart beat position of the ECG signal (n), which is represented by the ROC sequence; the kth heart beat is represented by ROC (k);
step three, on the kth heartbeat ROC (k), respectively intercepting ECG fragments from ROC (k) to ROC (k) in the first 180ms and fragments from ROC (k) to ROC (k) in the last 220ms, and respectively recording the ECG fragments as sig1 and sig 2; extracting feature points of sig1 and sig2 by a Douglas-Peucker algorithm; respectively assuming that the feature points extracted from sig1 and sig2 are P and Q, and respectively recording the feature points into vertex1 and vertex 2; because the start point, the stop point, the Q wave peak and the S wave peak of the QRS wave belong to the category of the feature points, the vertex1 comprises the information of the start point and the Q wave peak of the QRS wave, and the vertex2 comprises the information of the stop point and the S wave peak of the QRS wave;
step four, screening candidate points of the QRS wave starting point and the QRS wave ending point from vertex1 and vertex2 respectively, and recording Q _ points _ candidate and S _ points _ candidate respectively;
step five, sorting QRS wave candidate starting points in the Q _ points _ candidate according to the time sequence of the QRS wave candidate starting points in the sig1, taking the candidate point sorted at the first position as the starting point of the QRS wave, and recording the Qenset; similarly, sorting QRS wave candidate termination points in the S _ points _ candidate according to the time sequence of the QRS wave candidate termination points in the sig2, taking the candidate point sorted at the last position as the termination point of the QRS wave, and recording the QRS wave candidate termination point in Soffset; QRS wave width is the difference between Soffset and Qoffset;
and step six, performing the same processing on each heart beat in signal (n) to position the start point and the stop point of the QRS wave of each heart beat.
Further, in the step 4: for vertex1, limiting the number of candidate start points of the QRS wave to be screened to limit _ hum, and the specific steps are as follows:
step a: selecting an optional point from vertex1 as an initial reference point, and recording as ref _ point; comparing a characteristic point nearest to the ref _ point with the ref _ point, if the following 3 conditions are met, the compared characteristic point can be used as an effective QRS wave candidate starting point and stored into a Q _ points _ candidate, and the point is used as a new ref _ point; otherwise, abandoning the characteristic point, reselecting another characteristic point which is closest to the ref _ point as a comparison point, and repeating the comparison process;
an amplitude condition; the amplitude difference of the compared characteristic point and ref _ point in sig1 is greater than an amplitude threshold h _ sh;
a time difference condition; the time difference between the compared characteristic point and the position of ref _ point in sig1 is greater than a threshold value t _ sh;
(ii) a relief condition; the compared characteristic points are opposite to the concavity and convexity of ref _ point in sig 1; if the concave-convex characteristics of the two are the same, the absolute value of the amplitude of the compared characteristic point in the sig1 is required to be larger than the absolute value of the amplitude of ref _ point;
step b: repeating the step a with the new ref _ point until all the feature points in vertex1 are compared; if more than 2 continuous comparison feature points can not meet 3 conditions at the same time, all QRS wave candidate starting points are found from vertex1, the comparison is terminated, and the next step is directly carried out;
step c: if the QRS wave candidate starting point number recorded in the Q _ points _ candidate is less than or equal to the limit _ num, stopping screening; otherwise, increasing the amplitude threshold h _ sh, and repeating the steps a and b;
when selecting the candidate QRS termination point from vertex2, after steps a-c are performed, the candidate termination point needs to be retested to determine the final candidate QRS termination point, and then the final candidate QRS termination point is recorded in S _ points _ candidate.
The invention has the beneficial effects that:
(1) compared with the prior art, the R wave (namely ECG heart beat) positioning scheme provided by the invention has higher positioning accuracy under the conditions of stronger noise and abnormal QRS wave form.
(2) The method can keep higher QRS wave start and stop point positioning accuracy under the condition of abnormal QRS wave form.
(3) The method has certain identification capability on the occurrence of R 'waves or S' waves in the QRS wave complex.
(4) The method can solve the problems of the prior art that the QRS complex heart beat positioning accuracy rate is reduced and the QRS complex start and stop point positioning accuracy is poor under the conditions of ECG form change and high noise.
(5) By way of example, the more complex the QRS complex morphology, the more QRS complex candidate starting and ending points are retained by the present invention. Other candidate points provide information on the Q peak and S peak positions, except for the start point and the end point. Especially when multiple peaks such as R 'and S' appear in the QRS complex, the invention can also identify and locate the QRS complex. This advantage of the present invention is not available with other methods.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the embodiments or the drawings needed to be practical in the prior art description, and obviously, the drawings in the following description are only some embodiments of the embodiments, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
FIG. 1 is a flow chart of the method of the present invention.
FIG. 2 is a schematic illustration of the heart beat positioning process of the present invention;
FIG. 3 is a schematic diagram of the heart beat positioning result of the present invention;
fig. 4 is a diagram illustrating the effect of locating the start and stop points of the QRS complex according to the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the technical solutions of the present invention will be described in detail below. It is to be understood that the described embodiments are merely exemplary of the invention, and not restrictive of the full scope of the invention. All other embodiments, which can be derived by a person skilled in the art from the examples given herein without any inventive step, are within the scope of the present invention.
In any embodiment, as shown in fig. 1, a method for positioning a start point and a stop point of a QRS wave of an electrocardiograph signal according to the present invention includes the following steps:
step one, preprocessing an ECG signal;
the pretreatment comprises the following steps: filtering out baseline wander in the ECG by adopting a median filtering technology; to reduce the effect of floating point operations, the filtered ECG is scaled 500 times, denoted by signal (n); wherein n represents a sampling point; for an ECG signal of length N, then N is 1,2, …, N;
step two, detecting the ECG heartbeat position; piecewise fitting the signal (n) by using a least square quadratic polynomial technology, and reconstructing the signal (n) by using the slope and amplitude obtained by fitting so as to enhance the anti-noise capability of the signal; the reconstructed signal (n) is represented by s1(n), and the sampling points of the two samples are in one-to-one correspondence; extracting sampling points corresponding to signals with the amplitude larger than a threshold value from s1(n), and recording the sampling points into an R _ candidate sequence; calculating the difference of the R _ candidate sequence to obtain a difference sequence DR _ candidate; identifying from the DR _ candidate sequence that the differential value is greater than 0.25 fsThe sampling point of (1) and its immediately preceding sampling point; detecting the amplitude maximum value of s1(n) between all two sampling point pairs, and detecting the sampling point corresponding to the maximum value as the R peak position, namely the heart beat position of the ECG signal (n), which is represented by the ROC sequence; RO for kth heart beatC (k) represents; the method comprises the following specific steps:
step 1: a sliding window of width 40ms is set according to the cardiac periodicity of the ECG signal, with the position of the window being represented by the position of the center of the window. The center position of the window is first aligned to the first sample point of signal (n) (zero padding where there is no ECG signal within the window), and then the window is slid with 1 sample point as the step size until the last sample point. Therefore, the position k of the sliding window corresponds to the sampling point N of signal (N) in a one-to-one manner, where k is 1,2, … …, N. The ECG segments within each window are fitted with a least squares 2 degree polynomial of equation (1) to obtain the slope cur (k) and amplitude amp (k) characterizing each ECG segment (corresponding to the 2 nd order coefficients and constants in the 2 nd order polynomial (1), respectively). In order to guarantee the accuracy of fitting, the least square fitting error is set to be less than 0.001.
signal(n)≈-cur(n-k)2+amp (1)
In equation (1), amp is the amplitude parameter (amplitude) obtained at the n point in the signal ECG, cur is the slope parameter (current) obtained.
Step 2: using the slope cur (k) and amplitude amp (k) of the ECG segment in the kth window to reconstruct a signal as in equation (2), denoted s1(k)
s1(n)=ws(n)·*0.6amp(n)·*0.4cur(n) (2)
Wherein, ws(k) Is a weighting coefficient, and is calculated according to equation (3).
And step 3: and step 3: reconstructed signals satisfying the condition of equation (4) are extracted from s1(k), and it is assumed that N1 such reconstructed signals are extracted.
s1(k)>s1(k-1),s1(k)>s1(k+1) (4)
Half of the average of the N1 reconstructed signals is taken as the threshold. Since the QRS wave has the largest amplitude in the ECG, the segment of the signal with amplitude greater than the threshold at s1(k) usually corresponds to the QRS complex segment in the ECG, and for this reason, the sampling points of this part of the signal are recorded in the R _ candidate sequence.
And 4, step 4: the first order difference is calculated for the R _ candidate sequence, and the difference result is entered into the DR _ candidate sequence. From the rhythmicity of the ECG, the R peak is located between the sample point with a differential value greater than 0.25 x fs and the immediately preceding sample point. As shown in fig. 2. The signal in the figure is an auxiliary signal s1, and triangles and circles respectively represent a sampling point with a difference value larger than 0.25 x fs and adjacent sampling points, and a convex peak part between two adjacent points represents the position of an R peak. The value 0.25 is because the heart rate is estimated to be no more than 200bpm at the fastest, that is, no more than 4 heartbeats per second, that is, 1/4 is 0.25. The amplitude maximum of s1(n) is detected between all two pairs of samples in the DR _ candidate sequence, and its corresponding sample is detected as the R peak position, i.e., the heart beat position of the ECG signal, represented by the ROC sequence. The kth heart beat is denoted by ROC (k), and the result is shown in FIG. 3, where the circle is located, i.e., the R peak-to-peak point.
Step three, on the kth heartbeat ROC (k), respectively intercepting ECG fragments from ROC (k) to ROC (k) in the first 180ms and fragments from ROC (k) to ROC (k) in the last 220ms, and respectively recording the ECG fragments as sig1 and sig 2; extracting feature points of sig1 and sig2 by a Douglas-Peucker algorithm; respectively assuming that the feature points extracted from sig1 and sig2 are P and Q, and respectively recording the feature points into vertex1 and vertex 2; because the start point, the stop point, the Q wave peak and the S wave peak of the QRS wave belong to the category of the feature points, the vertex1 comprises the information of the start point and the Q wave peak of the QRS wave, and the vertex2 comprises the information of the stop point and the S wave peak of the QRS wave;
the QRS wave candidate starting point screening method specifically comprises the following steps:
step 1: intercepting signal segments sig1 from ROC (k) to ROC (k) in electrocardiosignals in the first 180ms, and extracting 10 characteristic points from sig1 by adopting a Douglas-Peucker algorithm. The reason why P is 10 feature points is that it is found that the QRS complex starting point is most accurately located when P is 10 during the test. First, the head and tail points A and B of the curve sig1 are connected to obtain a chord AB of the curve sig 1. Searching for distances from curve sig1Point C furthest from chord AB1And C is1As a feature point on sig 1. C1Point divides sig1 into two curves, followed by connection of AC1And BC1Then AC1And BC1Respectively the corresponding chords of the two curves. Respectively searching distance chord AC from two sections of curves1And BC1The two farthest points are calculated, the farthest distance is calculated, and the larger of the two points is taken as the second characteristic point C in the sig12. Characteristic point C1And C2Dividing sig1 into three segments of AC1、C1C2And BC2Similarly, the point farthest from the three-segment chord is found on sig1 and the distance value from the chord is calculated, and the point with the largest distance value is taken as the third characteristic point C3. And so on until all 10 feature points are extracted and logged into vertex 1.
Step 2: the amplitude condition h _ sh1, the time difference condition t _ sh1, the concavity and convexity condition direction1, and the candidate starting point number limit condition limit _ num1 are calculated respectively by using equations (5) to (7). And screening the QRS wave candidate starting points from the 10 characteristic points according to the conditions.
h_sh1=h_sh0*500 (5)
In equation (5), h _ sh0 represents the initial amplitude, which is an empirical value and is set to 0.05 mv.
In equation (6), dh is the absolute value of the difference between the compared point and the amplitude value of the reference point ref _ point on sig 1. []In order to round the symbol, the symbol is rounded,to round down the symbols, each constant value in the equation is an empirical value.
In the formula (7), vertex1(i) represents the ith comparison point in chronological order in vertex 1; sig1(vertex1(i)) is the amplitude value of the ith alignment point in sig1, direction1 is equal to 1, which means that the ECG segment in which the alignment point is located is upwardly convex, and direction1 is equal to-1, and is downwardly concave.
In equation (8), sig1(1) represents the amplitude at the first sampling point in the ECG signal segment sig 1.
And step 3: and if the QRS wave candidate starting point number in the Q _ points _ candidate is greater than the limit _ num1, increasing the initial amplitude according to the formula (9) and repeating the step 2. Otherwise, finishing the screening of the QRS wave candidate starting point.
h_sh0=h_sh0+0.005 (9)
The QRS wave candidate termination point screening method specifically comprises the following steps:
step 1: and intercepting signal segments within the range from ROC (k) to ROC (k) within 220ms, recording the signal segments as sig2, and extracting 6 feature points from sig2 by adopting a Douglas-Peucker algorithm. The reason for taking Q6 feature points here is that during the test, it was found that locating the QRS complex endpoint was most accurate when Q6. The implementation is exactly the same as extracting feature points from sig 1. The 6 feature points extracted from sig2 were entered into vertex 2.
Step 2: the amplitude condition h _ sh2, the time difference condition t _ sh2, the irregularity condition direction2, and the candidate start point number limit condition limit _ num2 are calculated using the formulas (10), (11), (13), and (14), respectively. And screening the QRS wave candidate terminal points from the 6 characteristic points according to the conditions.
In the formula (10), i represents the ith feature point used for comparison after being sorted in time sequence in vertex2, and ref _ min _ point is the position of the feature point with the minimum amplitude in vertex 2.
In the formula (11), dh is the absolute value of the amplitude difference between the compared point and the reference point, and t _ sh0 is the initial threshold of the time difference and is taken according to table 1. w is a weight variable relating to the amplitude difference, and is calculated from equation (12), and the constant 0.056 is an empirical value.
TABLE 1 t _ sh0Value of
In table 1 and equation (12), the direction2(1) is the concavity and convexity of the first point in vertex2 sorted in time series.
In the formula (13), vertex2(i) represents the position of the ith comparison point in vertex2 in the signal, where i has the same meaning as above. sig2(vertex2(i)) is the magnitude value of the ith alignment point in sig 2. The direction2 has the same meaning as the direction1 described above.
In equation (14), sig2(1) is the amplitude value of the first sample point in sig 2.
And step 3: and if the QRS wave candidate termination point number in the S _ points _ candidate is more than the limit _ num2, increasing the initial amplitude according to the formula (9) and repeating the step 2. Otherwise, go to step 4.
And 4, step 4: if one of the following two conditions occurs in the QRS wave candidate termination point number in the S _ points _ candidate, the feature point needs to be rechecked. Otherwise, finishing the screening of the QRS wave candidate termination point.
In the first case: the QRS wave candidate termination point number in the S _ points _ candidate is 1, and the QRS wave candidate termination point number is smaller than-0.3 mv, wherein the-0.3 mv is an empirical value. In this case, an ECG signal segment from the candidate point position to the end of sig2 is intercepted from sig2, 6 new feature points are extracted from the ECG segment by adopting a Douglas-Peucker algorithm, and the new feature points are rechecked to obtain a true QRS wave candidate termination point.
In the second case: the QRS wave candidate termination point in the S _ points _ candidate is 0. In this case, the recheck is directly performed on the feature point set in vertex2, and the true QRS wave candidate end point is obtained.
The retest embodiment is the same for both cases. The second case is taken as an example to describe the reinspection embodiment.
(1) Firstly, according to the formulas (15) and (16), respectively calculating the slope s (i) of the ith characteristic point which is used for comparison and is sorted in time sequence in vertex2 and the Euclidean distance ds (i) between the ith characteristic point and i-1 characteristic point;
normalized Euclidean distance ds (i) is shown in formula (17)
(2) The score of the ith feature point is calculated using equation (18).
In the formula (18), s (i) represents the slope calculated by the formula (17), score (i) represents the score of the i-th feature point, and direction (i) represents the concavity and convexity of the i-th feature point. sin () represents a sine function and arctan () is an arctangent function. Here flag takes the values according to Table 2.
TABLE 2 flag values and their values
flag | |
1 | mbuff_sig2_1-mbuff_sig2-2>0.2*500 |
1 | mbuff_sig2_2-mbuff_sig2_1<-0.1*500&count_d≥6 |
1 | mdbuff_sig2_1<0&count_d≥6 |
0 | Others |
Mbuff _ sig2_1 and mbuff _ sig2_2 in table 2 represent the average of the amplitudes of the first 8 samples and the last 8 samples, respectively, in signal segment sig 2. mdmuff _ sig2_1 and count _ d represent the average value of the amplitude values after 1-order difference of the first 8 sampling points in sig2 and the number of the difference values smaller than 0.
Step four, screening candidate points of the QRS wave starting point and the QRS wave ending point from vertex1 and vertex2 respectively, and recording Q _ points _ candidate and S _ points _ candidate respectively;
in the fourth step: for vertex1, limiting the number of candidate start points of the QRS wave to be screened to limit _ num, and the specific steps are as follows:
step a: selecting an optional point from vertex1 as an initial reference point, and recording as ref _ point; comparing a characteristic point nearest to the ref _ point with the ref _ point, if the following 3 conditions are met, the compared characteristic point can be used as an effective QRS wave candidate starting point and stored into a Q _ points _ candidate, and the point is used as a new ref _ point; otherwise, abandoning the characteristic point, reselecting another characteristic point which is closest to the ref _ point as a comparison point, and repeating the comparison process;
an amplitude condition; the amplitude difference of the compared characteristic point and ref _ point in sig1 is greater than an amplitude threshold h _ sh;
a time difference condition; the time difference between the compared characteristic point and the position of ref _ point in sig1 is greater than a threshold value t _ sh;
(ii) a relief condition; the compared characteristic points are opposite to the concavity and convexity of ref _ point in sig 1; if the concave-convex characteristics of the two are the same, the absolute value of the amplitude of the compared characteristic point in the sig1 is required to be larger than the absolute value of the amplitude of ref _ point;
step b: repeating the step a with the new ref _ point until all the feature points in vertex1 are compared; if more than 2 continuous comparison feature points can not meet 3 conditions at the same time, all QRS wave candidate starting points are found from vertex1, the comparison is terminated, and the next step is directly carried out;
step c: if the QRS wave candidate starting point number recorded in the Q _ points _ candidate is less than or equal to the limit _ num, stopping screening; otherwise, increasing the amplitude threshold h _ sh, and repeating the steps a and b;
when selecting the candidate QRS termination point from vertex2, after steps a-c are performed, the candidate termination point needs to be retested to determine the final candidate QRS termination point, and then the final candidate QRS termination point is recorded in S _ points _ candidate.
Step five, sorting QRS wave candidate starting points in the Q _ points _ candidate according to the time sequence of the QRS wave candidate starting points in the sig1, taking the candidate point sorted at the first position as the starting point of the QRS wave, and recording the Qenset; similarly, sorting QRS wave candidate termination points in the S _ points _ candidate according to the time sequence of the QRS wave candidate termination points in the sig2, taking the candidate point sorted at the last position as the termination point of the QRS wave, and recording the QRS wave candidate termination point in Soffset; QRS wave width is the difference between Soffset and Qoffset;
and step six, performing the same processing on each heart beat in signal (n) to position the start point and the stop point of the QRS wave of each heart beat.
The QRS wave heart beat detection and start and stop point positioning experiment is carried out by applying the invention.
The test data was from the qtdb (qt database) database, totaling 105 groups, all manually labeled by experts with QRS complex R peak and start and stop point information. Each set of data consists of two leads, 15 minutes in duration, with a sampling rate of 250 Hz. This example only selects the first lead ECG data for testing the ability of the present invention to process single lead ECG data.
This example is implemented on a Personal Computer (PC). The technical indexes of the personal computer are as follows: the processor is Intel (R) core (TM) i 56500, the memory is 8GB, the operating system is window7, and the programming software is Matlab 2014 a.
On QRS wave heartbeat detection, if the error between the detection result and the expert mark is within 100ms, the detection result is regarded as correct detection, which is strictly under the internationally accepted criterion ANSI/AAMI EC 38. The method adopts Sensitivity (SE) and positive prediction rate (P +) which can comprehensively reflect the detection performance to evaluate the performance of the method, and is specifically shown as formulas (21) and (22).
True Positives (TP) indicate the number of correct detections, False Positives (FP) indicate the number of false detections, and false negatives FN indicate the number of missed detections. Meanwhile, the detection method of the invention is compared with other related methods in performance. The results obtained on the test data for the present invention and other related methods are shown in table 3. Fig. 3 is a graph showing the effect of detecting the heart beat of the QRS wave in the invention.
TABLE 3 comparison of the Performance of the present invention in detecting ECG heartbeats with other related algorithms
Class of algorithms | Sensitivity SE | Positive prediction rate P + |
Pan and Tompkins | 0.9893 | 0.9904 |
Manriquez A I et al. | 0.9958 | 0.9958 |
Antoni Burguera | 0.9784 | 0.9784 |
The method of the invention | 0.9983 | 0.9985 |
In the positioning of the QRS wave group start and stop points, the invention adopts the mean value and the variance of the error of the position of the start and stop points marked by experts to evaluate the performance of the positioning method and compare the performance with other related methods. The results obtained on the test data for the present invention and other related methods are shown in table 4. Fig. 4 shows the effect of locating the start and stop points of the QRS complex according to the present invention, wherein three circle positions from left to right represent the start point, the R peak point and the end point of the located QRS complex, respectively.
TABLE 4 comparison of performance of the present invention and other related algorithms for locating the onset and termination points of QRS complexes
As can be seen from Table 3, the sensitivity and the positive prediction rate of the QRS complex in the heart beat detection of the QRS complex of the invention reach 99.83% and 99.85%, respectively, which are superior to all comparison methods. As can be seen from Table 4, compared with other related methods, the method of the present invention has a slight advantage in positioning the QRS complex start point and end point in the standard deviation, and has a significant advantage in the mean error.
In order to verify the detection capability of the method when the ECG morphology changes, typical electromyographic noise and baseline drift noise provided by MIT-NSTDB are added into ECG data, the interference on the ECG is simulated and collected under the real condition, three groups of ECG data with signal-to-noise ratios of 24dB, 12dB and 0dB are obtained, the three groups of ECG data are respectively used for verifying the performance of the method for detecting the QRS wave group starting and stopping points under the condition that the ECG morphology changes, and the results are compared with other related methods, and are shown in Table 5.
TABLE 5 comparison of the performance of this patent in detecting the start and stop points of QRS complex with other methods under different SNR
As shown in table 5, the identification accuracy of the QRS complex at both the starting point and the ending point is better than that of the alignment method at the snr of 24dB, 12dB and 0dB, respectively. Especially in case of severe noise (0dB) interference, the overall mean error is less than 8ms and the standard error is less than 20.4 ms. Calculated at a sampling frequency of 250hz, the error in the tracing of the real electrocardiogram does not exceed 0.5mm (20ms), which is only similar to the line width of the tracing waveform. Therefore, the capability of detecting the QRS wave group start and stop points is stronger than that of other related methods under the condition that the ECG waveform is distorted due to noise interference, and the obtained result has higher reliability.
It is worth noting that in the example, we find that the more complex the QRS complex morphology is, the more candidate starting and ending points of the QRS complex are retained by the present invention. Other candidate points provide information on the Q peak and S peak positions, except for the start point and the end point. Especially when multiple peaks such as R 'and S' appear in the QRS complex, the invention can also identify and locate the QRS complex. This advantage of the present invention is not available with other methods.
The above description is only for the specific embodiments of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present invention, and all the changes or substitutions should be covered within the scope of the present invention. The various features described in the foregoing detailed description may be combined in any suitable manner without contradiction, and various combinations that are possible in the present invention will not be further described in order to avoid unnecessary repetition. Any combination of the different embodiments of the present invention is also possible, and the same should be considered as the disclosure of the present invention as long as it does not depart from the gist of the present invention.
Claims (4)
1. A method for positioning a start point and a stop point of a QRS wave of an electrocardiosignal is characterized by comprising the following steps:
step one, preprocessing an ECG signal;
the pretreatment comprises the following steps: filtering out baseline wander in the ECG by adopting a median filtering technology; to reduce the effect of floating point operations, the filtered ECG is scaled 500 times, denoted by signal (n); wherein n represents a sampling point; for an ECG signal of length N, then N ═ 1,2, ·, N;
step two, detecting the ECG heartbeat position; piecewise fitting the signal (n) by using a least square quadratic polynomial technology, and reconstructing the signal (n) by using the slope and amplitude obtained by fitting so as to enhance the anti-noise capability of the signal; the reconstructed signal (n) is represented by s1(n), and the sampling points of the two samples are in one-to-one correspondence; extracting amplitude greater than threshold from s1(n)Sampling points corresponding to the signal of threshold, and recording an R _ candidate sequence; calculating the difference of the R _ candidate sequence to obtain a difference sequence DR _ candidate; identifying from the DR _ candidate sequence that the differential value is greater than 0.25 fsThe sampling point of (1) and its immediately preceding sampling point; detecting the amplitude maximum value of s1(n) between all two sampling point pairs, and detecting the sampling point corresponding to the maximum value as the R peak position, namely the heart beat position of the ECG signal (n), which is represented by the ROC sequence; the kth heart beat is represented by ROC (k);
step three, on the kth heartbeat ROC (k), respectively intercepting ECG fragments from ROC (k) to ROC (k) in the first 180ms and fragments from ROC (k) to ROC (k) in the last 220ms, and respectively recording the ECG fragments as sig1 and sig 2; extracting feature points of sig1 and sig2 by a Douglas-Peucker algorithm; respectively assuming that the feature points extracted from sig1 and sig2 are P and Q, and respectively recording the feature points into vertex1 and vertex 2; because the start point, the stop point, the Q wave peak and the S wave peak of the QRS wave belong to the category of the feature points, the vertex1 comprises the information of the start point and the Q wave peak of the QRS wave, and the vertex2 comprises the information of the stop point and the S wave peak of the QRS wave;
step four, screening candidate points of the QRS wave starting point and the QRS wave ending point from vertex1 and vertex2 respectively, and recording Q _ points _ candidate and S _ points _ candidate respectively; in the fourth step: for vertex1, limiting the maximum number of QRS wave candidate starting points to be screened to limit _ num, and the specific steps are as follows:
step a: selecting an optional point from vertex1 as an initial reference point, and recording as ref _ point; comparing a characteristic point nearest to the ref _ point with the ref _ point, if the following 3 conditions are met simultaneously, the characteristic point can be used as an effective QRS wave candidate starting point and stored into the Q _ points _ candidate, and the characteristic point is used as a new ref _ point; otherwise, abandoning the characteristic point, reselecting another characteristic point which is closest to the ref _ point as a comparison point, and repeating the comparison process;
an amplitude condition; the amplitude difference of the compared characteristic point and ref _ point in sig1 is greater than an amplitude threshold h _ sh;
a time difference condition; the time difference between the compared characteristic point and the position of ref _ point in sig1 is greater than a threshold value t _ sh;
(ii) a relief condition; the compared characteristic points are opposite to the concavity and convexity of ref _ point in sig 1; if the concave-convex characteristics of the two are the same, the absolute value of the amplitude of the compared characteristic point in the sig1 is required to be larger than the absolute value of the amplitude of ref _ point;
step b: repeating the step a by taking the comparison points meeting the 3 conditions as new ref _ point until all the feature points in vertex1 are compared; if more than 2 continuous comparison feature points can not meet 3 conditions at the same time, all QRS wave candidate starting points are found from vertex1, the comparison is terminated, and the next step is directly carried out;
step c: if the QRS wave candidate starting point number recorded in the Q _ points _ candidate is less than or equal to the limit _ num, stopping screening; otherwise, increasing the amplitude threshold h _ sh, and repeating the steps a and b;
when selecting the candidate QRS termination points from vertex2, after steps a-c are executed, the candidate termination points are required to be rechecked, the final candidate QRS termination points are determined, and S _ points _ candidate is recorded;
step five, sorting QRS wave candidate starting points in the Q _ points _ candidate according to the time sequence of the QRS wave candidate starting points in the sig1, taking the candidate point sorted at the first position as the starting point of the QRS wave, and recording the Qenset; similarly, sorting QRS wave candidate termination points in the S _ points _ candidate according to the time sequence of the QRS wave candidate termination points in the sig2, taking the candidate point sorted at the last position as the termination point of the QRS wave, and recording the QRS wave candidate termination point in Soffset; QRS wave width is the difference between Soffset and Qoffset;
and step six, performing the same processing on each heart beat in signal (n) to position the start point and the stop point of the QRS wave of each heart beat.
2. The method for positioning the start and stop point of the QRS wave of the electrocardiographic signal according to claim 1, wherein the second step is realized by the following steps:
step 1: setting a sliding window with the width of 40ms according to the cardiac periodicity of the ECG signal, and representing the position of the window by the central position of the window; firstly, aligning the central position of a window with the first sampling point of signal (n), and then sliding the window by taking 1 sampling point as a step length until the last sampling point; therefore, the position k of the sliding window corresponds to the sampling point N of the signal (N), where k is 1,2, · N; fitting the ECG segments in each window by adopting a least square 2-order polynomial of formula (1) to obtain a slope cur (k) and an amplitude amp (k) which represent each ECG segment; in order to guarantee the accuracy of fitting, the least square fitting error is set to be less than 0.001;
signal(n)≈-cur(n-k)2+amp (1)
in equation (1), amp is the amplitude parameter (amplitude) obtained at n points in the signal ECG, cur is the slope parameter (slope) obtained;
step 2: a signal is reconstructed using the slope cur (k) and amplitude amp (k) of the ECG segment in the kth window as shown by equation (2) s1 (k):
s1(k)=ws(k)·*0.6amp(k)·*0.4cur(k) (2)
wherein, ws(k) Is a weighting coefficient, calculated according to equation (3);
and step 3: extracting reconstructed signals satisfying the condition of equation (4) from s1(k), and assuming that N1 such reconstructed signals are extracted;
s1(k)>s1(k-1),s1(k)>s1(k+1) (4)
taking half of the average value of the N1 reconstructed signals as a threshold; since the QRS wave has the largest amplitude in the ECG, the segment of the signal with amplitude greater than the threshold in s1(k) usually corresponds to the QRS complex segment in the ECG, and for this reason, the sampling points of this part of the signal are recorded in the R _ candidate sequence;
and 4, step 4: calculating a first order difference of the R _ candidate sequence, and recording the difference result into the DR _ candidate sequence; from the rhythmicity of the ECG, the R peak is located between the sample point with a differential value greater than 0.25 x fs and the immediately preceding sample point.
3. The method for locating the start and stop points of the QRS wave of an electrocardiograph signal according to claim 1, wherein in the third step, the specific method for screening the candidate start points of the QRS wave is as follows:
step 1: intercepting signal segments sig1 from ROC (k) to ROC (k) in electrocardiosignals in the first 180ms, and extracting 10 characteristic points from sig1 by adopting a Douglas-Peucker algorithm; the reason why P is 10 feature points is that in the test process, it is found that the QRS complex starting point is located most accurately when P is 10; firstly, connecting a head point A and a tail point B of a curve sig1 to obtain a chord AB of a curve sig 1; from curve sig1, search is made for the point C farthest from chord AB1And C is1As a feature point on sig 1; c1Point divides sig1 into two curves, followed by connection of AC1And BC1Then AC1And BC1Respectively corresponding chords of the two sections of curves; respectively searching distance chord AC from two sections of curves1And BC1The two farthest points are calculated, the farthest distance is calculated, and the larger of the two points is taken as the second characteristic point C in the sig12(ii) a Characteristic point C1And C2Dividing sig1 into three segments of AC1、C1C2And BC2Similarly, the point farthest from the three-segment chord is found on sig1 and the distance value from the chord is calculated, and the point with the largest distance value is taken as the third characteristic point C3(ii) a And so on until all 10 feature points are extracted and recorded in vertex 1;
step 2: respectively calculating an amplitude condition h _ sh1, a time difference condition t _ sh1, a concavity and convexity condition direction1 and a candidate starting point limit condition limit _ num1 by using formulas (5) to (7); then according to the conditions, screening a QRS wave candidate starting point from the 10 characteristic points;
h_sh1=h_sh0*500 (5)
in the formula (5), h _ sh0 represents the initial amplitude, is an empirical value and is set to 0.05 mv;
in equation (6), dh is the absolute value of the difference between the comparison point and the amplitude value of the reference point ref _ point on sig 1;in order to round the symbol, the symbol is rounded,in order to get the sign of rounding down, each constant value in the formula is an empirical value;
in the formula (7), vertex1(i) represents the ith comparison point in chronological order in vertex 1; sig1(vertex1(i)) is the amplitude value of the ith comparison point in sig1, direction1 is equal to 1, which means that the ECG segment in which the comparison point is located is upward convex, and direction1 is equal to-1, and then downward concave;
in equation (8), sig1(1) represents the amplitude at the first sampling point in the ECG signal segment sig 1;
and step 3: if the QRS wave candidate starting point number in the Q _ points _ candidate is larger than the limit _ num1, increasing the initial amplitude according to the formula (9) and then repeating the step 2; otherwise, finishing the screening of the QRS wave candidate starting point;
h_sh0=h_sh0+0.005 (9)。
4. the method for locating the start and stop point of the QRS wave of an electrocardiograph signal according to claim 1, wherein in the third step, the step of screening the candidate start and stop points of the QRS wave comprises the following specific steps:
step 1: intercepting signal segments within a range from ROC (k) to ROC (k) within 220ms, recording the signal segments as sig2, and extracting 6 feature points from sig2 by adopting a Douglas-Peucker algorithm; the reason why the Q is 6 characteristic points is that the QRS complex endpoint is located most accurately when the Q is 6 in the test process; the implementation is exactly the same as the extraction of feature points from sig 1; recording 6 characteristic points extracted from sig2 into vertex 2;
step 2: respectively calculating an amplitude condition h _ sh2, a time difference condition t _ sh2, an irregularity condition direction2 and a candidate starting point limit condition limit _ num2 by adopting formulas (10), (11), (13) and (14); screening a QRS wave candidate terminal point from the 6 characteristic points according to the conditions;
in the formula (10), i represents the ith feature point used for comparison after being sorted in time sequence in vertex2, and ref _ min _ point is the feature point position with the minimum amplitude in vertex 2;
in the formula (11), dh is an absolute value of the amplitude difference between the point to be compared and the reference point, t _ sh0 is an initial threshold value of the time difference, w is a weight variable related to the amplitude difference, the weight variable is obtained by calculation according to the formula (12), and a constant 0.056 is an empirical value;
in the formula (12), the direction2(1) is the concavity and convexity of the first point in vertex2 after being sorted in time sequence;
in the formula (13), vertex2(i) represents the position of the ith comparison point in vertex2 in the signal, i represents the ith characteristic point which is used for comparison after being sorted in time sequence in vertex 2; sig2(vertex2(i)) is the magnitude value of the ith alignment point in sig 2; direction2 has the same meaning as direction1 described above;
in equation (14), sig2(1) is the amplitude value of the first sample point in sig 2;
and step 3: if the QRS wave candidate termination point number in the S _ points _ candidate is greater than the limit _ num2, increasing the initial amplitude according to the formula (9) and then repeating the step 2; otherwise, entering step 4;
and 4, step 4: if the QRS wave candidate termination point number in the S _ points _ candidate appears in one of the following two situations, the feature point needs to be rechecked; otherwise, completing the screening of the QRS wave candidate termination point;
in the first case: the QRS wave candidate termination point number in the S _ points _ candidate is 1, the amplitude of the QRS wave candidate termination point number is less than-0.3 mv, and the-0.3 mv is an empirical value; under the condition, an ECG signal segment from the candidate point position to the tail end of the sig2 is intercepted from the sig2, 6 new feature points are extracted from the ECG segment by adopting a Douglas-Peucker algorithm, and the new feature points are subjected to rechecking to obtain a true QRS wave candidate termination point;
in the second case: the QRS wave candidate termination point number in the S _ points _ candidate is 0; in this case, the recheck is directly performed on the feature point set in vertex2, and the true QRS wave candidate end point is obtained.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910793288.5A CN110477906B (en) | 2019-08-26 | 2019-08-26 | Electrocardiosignal QRS wave start and stop point positioning method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910793288.5A CN110477906B (en) | 2019-08-26 | 2019-08-26 | Electrocardiosignal QRS wave start and stop point positioning method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110477906A CN110477906A (en) | 2019-11-22 |
CN110477906B true CN110477906B (en) | 2021-11-26 |
Family
ID=68554341
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910793288.5A Active CN110477906B (en) | 2019-08-26 | 2019-08-26 | Electrocardiosignal QRS wave start and stop point positioning method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110477906B (en) |
Families Citing this family (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111134659B (en) * | 2020-01-17 | 2022-04-26 | 上海乐普云智科技股份有限公司 | Method and device for detecting P wave and T wave in electrocardiosignal |
CN111643070B (en) * | 2020-06-12 | 2023-05-23 | 京东方科技集团股份有限公司 | T wave starting point determining method and device, storage medium and electronic equipment |
CN112883803B (en) * | 2021-01-20 | 2023-09-01 | 武汉中旗生物医疗电子有限公司 | Electrocardiogram signal classification method, device and storage medium based on deep learning |
CN113317796A (en) * | 2021-06-22 | 2021-08-31 | 郑州大学 | Automatic electrocardio diagnosis method based on intelligent simulation modeling |
CN113712569B (en) * | 2021-11-01 | 2022-02-08 | 毕胜普生物科技有限公司 | High-frequency QRS wave group data analysis method and device |
CN114742114B (en) * | 2022-06-09 | 2022-10-21 | 毕胜普生物科技有限公司 | High-frequency QRS waveform curve analysis method and device, computer equipment and storage medium |
CN117357130B (en) * | 2023-12-07 | 2024-02-13 | 深圳泰康医疗设备有限公司 | Electrocardiogram digital curve segmentation method based on artificial intelligence |
CN118078305B (en) * | 2024-04-26 | 2024-06-21 | 深圳市泽辉医疗技术有限公司 | Dynamic electrocardiograph data processing method for wearable blood flow dynamic monitor |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103083013A (en) * | 2013-01-18 | 2013-05-08 | 哈尔滨工业大学深圳研究生院 | Electrocardio signal QRS complex wave detection method based on morphology and wavelet transform |
CN106197440A (en) * | 2016-06-27 | 2016-12-07 | 江苏迪纳数字科技股份有限公司 | Road data compression method based on angle restriction |
CN108294745A (en) * | 2018-03-07 | 2018-07-20 | 武汉大学 | P waves, T wave start-stop point detecting methods and system in multi-lead ECG signal |
Family Cites Families (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5033473A (en) * | 1989-10-24 | 1991-07-23 | Hewlett-Packard Company | Method for discriminating pace pulse tails |
US6560531B1 (en) * | 2001-05-22 | 2003-05-06 | Navigation Technologies Corporation | Multi-resolution trend metric for shape comparison and applications thereof |
EP2194857B1 (en) * | 2007-08-23 | 2011-11-30 | Cameron Health, Inc. | Patient screening tools for implantable cardiac stimulus systems |
EP2803141B1 (en) * | 2012-03-23 | 2017-11-01 | Siemens Aktiengesellschaft | Method and device for compressing data representing a time dependent signal |
EP2710953B1 (en) * | 2012-09-21 | 2015-03-04 | BIOTRONIK SE & Co. KG | Method of enhancing the signal-to-noise ratio (SNR) of measured electrocardiogram (ECG) signals and a cardiac device for use in detecting heartbeats |
CN104954094B (en) * | 2014-03-28 | 2018-01-23 | 华为技术有限公司 | Signal limiter method and apparatus |
CN104161510B (en) * | 2014-08-27 | 2016-03-23 | 电子科技大学 | A kind of multistage electrocardiosignal QRS waveform recognition methods of leading |
CN106725450B (en) * | 2017-02-20 | 2020-03-27 | 武汉中旗生物医疗电子有限公司 | QRS complex detection method |
CN109745035B (en) * | 2019-01-23 | 2021-04-20 | 深圳大学 | Electrocardiosignal waveform detection method |
CN109875550B (en) * | 2019-04-02 | 2020-08-04 | 东北大学 | Ventricular depolarization key point detection method |
-
2019
- 2019-08-26 CN CN201910793288.5A patent/CN110477906B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103083013A (en) * | 2013-01-18 | 2013-05-08 | 哈尔滨工业大学深圳研究生院 | Electrocardio signal QRS complex wave detection method based on morphology and wavelet transform |
CN106197440A (en) * | 2016-06-27 | 2016-12-07 | 江苏迪纳数字科技股份有限公司 | Road data compression method based on angle restriction |
CN108294745A (en) * | 2018-03-07 | 2018-07-20 | 武汉大学 | P waves, T wave start-stop point detecting methods and system in multi-lead ECG signal |
Also Published As
Publication number | Publication date |
---|---|
CN110477906A (en) | 2019-11-22 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110477906B (en) | Electrocardiosignal QRS wave start and stop point positioning method | |
CN110013247B (en) | Detection, differentiation and positioning method for electrocardiogram P wave and T wave | |
Kathirvel et al. | An efficient R-peak detection based on new nonlinear transformation and first-order Gaussian differentiator | |
CN101856225B (en) | Method for detecting R wave crest of electrocardiosignal | |
KR101524226B1 (en) | Method of classifing heart condition using neural network, and apparatus thereof | |
Campolo et al. | ECG-derived respiratory signal using empirical mode decomposition | |
Remya et al. | Classification of myocardial infarction using multi resolution wavelet analysis of ECG | |
Kumar et al. | Detection and classification of ECG noises using decomposition on mixed codebook for quality analysis | |
Choudhary et al. | A novel unified framework for noise-robust ECG-based biometric authentication | |
CN108992053B (en) | Method for real-time non-binding detection of heart rate and heartbeat interval | |
Singh et al. | Biometrics method for human identification using electrocardiogram | |
Manriquez et al. | An algorithm for QRS onset and offset detection in single lead electrocardiogram records | |
CN115486855B (en) | Electrocardiogram and heart beat classification method based on QRS complex indefinite cycle superposition | |
Sedighian et al. | Pediatric heart sound segmentation using Hidden Markov Model | |
CN111317499A (en) | Heart sound signal processing method based on wavelet technology | |
Manikandan et al. | Automatic SWT based QRS detection using weighted subbands and Shannon energy peak amplification for ECG signal analysis devices | |
Zhou et al. | Embedded real-time QRS detection algorithm for pervasive cardiac care system | |
Modak et al. | Single channel QRS detection using wavelet and median denoising with adaptive multilevel thresholding | |
Reklewski et al. | Real time ECG R-peak detection by extremum sampling | |
Varghees et al. | Two-channel heart sound segmentation framework using phonocardiogram and pulsatile signals | |
Illanes-Manriquez | An automatic multi-lead electrocardiogram segmentation algorithm based on abrupt change detection | |
CN105050493A (en) | Apparatus and method for determining the occurrence of a QRS complex in ECG data | |
JP5933138B2 (en) | Apparatus and method for determining occurrence of QRS complex in ECG data | |
Smrcka et al. | New methods for precise detection of systolic complexes in the signal acquired from quantitative seismocardiograph | |
Ruiz et al. | Thorough assessment of a p-wave delineation algorithm through the use of diverse electrocardiographic databases |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |