CN110169767B - Retrieval method of electrocardiosignals - Google Patents
Retrieval method of electrocardiosignals Download PDFInfo
- Publication number
- CN110169767B CN110169767B CN201910611770.2A CN201910611770A CN110169767B CN 110169767 B CN110169767 B CN 110169767B CN 201910611770 A CN201910611770 A CN 201910611770A CN 110169767 B CN110169767 B CN 110169767B
- Authority
- CN
- China
- Prior art keywords
- data
- retrieval
- retrieved
- electrocardiosignal
- calculating
- 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
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]
-
- 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
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/725—Details of waveform analysis using specific filters therefor, e.g. Kalman or adaptive filters
Abstract
The invention relates to a retrieval method of electrocardiosignals, which comprises the following steps: a. calculating the characteristics of the target data and the data to be retrieved; b. improving the electrocardiosignal similarity measurement of the DTW based on the TBD; c. evaluating the retrieval result by calculating the retrieval sensitivity and the positive retrieval rate; d. retrieving amplitude characteristics of the electrocardiogram data; e. morphological features of the electrocardiographic data are retrieved. The retrieval of the invention is mainly divided into the retrieval of the amplitude characteristic of the electrocardio data and the retrieval of the morphological characteristic of the electrocardio data. By processing by the method, the problems of electrocardiosignal similarity measurement are solved by fully utilizing the advantages of the electrocardiosignal and the electrocardiosignal, and retrieval of the electrocardiosignal is realized.
Description
Technical Field
The invention relates to an automatic detection and analysis technology of electrocardiosignals, in particular to a retrieval method of electrocardiosignals.
Background
At present, the diagnosis of cardiovascular diseases by doctors is mainly based on the judgment of the doctors on the electrocardiogram of patients. However, the cardiovascular diseases are of various types and have strong individual difference, so that the electrocardiogram of different patients shows different manifestations of the same disease. It is therefore not reasonable to use the same template for all individuals, which is also a clinical problem faced by automated diagnosis. Now, the workload of doctors is increased by massive electrocardiograms generated in clinical and telemedicine, so that the problem of how to assist the doctors to find interested electrocardio data in real time and quickly becomes an urgent need to be solved.
In order to greatly save the time of doctors and improve the diagnosis efficiency, the retrieval of the electrocardiosignal data is very necessary.
Disclosure of Invention
The invention aims to provide an electrocardiosignal retrieval method, which aims to solve the problems of poor retrieval precision of electrocardiosignal characteristic waves, poor generalization capability of manually designed characteristics and large characteristic dimension and more redundant information when the conventional retrieval method is used for similarity measurement.
The invention is realized by the following steps: a retrieval method of electrocardiosignals comprises the following steps:
a. calculating the characteristics of the target data and the data to be retrieved;
b. improving the electrocardiosignal similarity measurement of the DTW based on the TBD;
c. evaluating the retrieval result by calculating the retrieval sensitivity and the positive retrieval rate;
d. retrieving amplitude characteristics of the electrocardiogram data;
e. morphological features of the electrocardiographic data are retrieved.
The specific process of the step a is as follows:
a-1, determining target data a with length l ═ a1,a2,...al]Wherein, the amplitude feature vector of a is itself;
a-2, intercepting data to be retrieved b on the electrocardiosignal X according to the length l of the target data a1,b2,...bmWhen intercepting the data to be retrieved, selecting the step length as k, and calculating the expression of the data to be retrieved as follows:
data b to be retrieved1,b2,...bmThe amplitude feature vector is itself;
a-3, the expression of the morphological feature vector of the target data is
Wherein k is a positive integer, and the expression of the morphological feature vector of the data to be retrieved is
Wherein k is a positive integer.
The specific process of the step b is as follows:
b-1, calculating the distance of the points on the two sequences by adopting TBD: selecting target data Q ═ Q1,q2,...qm) Calculating data C ═ C1,c2,...cn) Distance from each point of Q, TBD is defined as
Then bring Q and C into
Wherein k is a positive integer, and a distance matrix D of m x n is obtained;
b-2, finding the optimal path from the point S to the point E in the distance matrix network, wherein the accumulated distance r (i, j) of each point in the distance matrix D is the sum of the accumulated distance r (i, j) and the shortest distance reaching the point, and the expression is as follows: r (i, j) ═ dTBD(qi,cj)+min(r(i-1,j-1),r(i-1,j),r(i,j-1))；
b-3, calculating the shortest distance between the selected data to be retrieved and the target data by adopting a method of improving DTW (dynamic time warping) through TBD (tunnel boring machine) to represent the similarity between the two data, wherein the similarity definition formula of the improved algorithm is as follows
The specific process of the step c is as follows: calculating the sensitivity and positive detection rate of the retrieval:
the sensitivity calculation formula is:
the positive detection rate is calculated by the formula:
wherein TP is the number of true positive samples, FN is the number of false negative samples, and FP is the number of false positive samples.
In the step d: the retrieval of the amplitude characteristics of the electrocardio data comprises the retrieval of the single-heart-beat electrocardio signal amplitude characteristics and the retrieval of the electrocardio signal amplitude characteristics with any length.
In the step e: the retrieval of the morphological characteristics of the electrocardio data comprises the retrieval of the morphological characteristics of the electrocardio signals of single heart beat and the retrieval of the morphological characteristics of the electrocardio signals with any length.
The invention discloses a retrieval method for improving dynamic time warping based on total Brazimann divergence, which mainly comprises two modes: A) retrieving amplitude characteristics of the electrocardio data; B) and (5) retrieving morphology features of the electrocardio data. By processing by the method, the problems of electrocardiosignal similarity measurement are solved by fully utilizing the advantages of the electrocardiosignal and the electrocardiosignal, and retrieval of the electrocardiosignal is realized.
Drawings
FIG. 1 is a flow chart of a single-beat amplitude retrieval algorithm.
FIG. 2 is a flow chart of a single-beat morphology retrieval algorithm.
Fig. 3 is a schematic view taken by heart.
FIG. 4 is a flow chart of an arbitrary length electrocardiosignal amplitude retrieval algorithm.
FIG. 5 is a flow chart of an arbitrary length electrocardiosignal morphology retrieval algorithm.
Detailed Description
The present invention is further described in detail below with reference to the attached drawings, and those skilled in the art can implement the present invention from the disclosure of the present specification.
The embodiment is realized in a computer with an Intel Xeon CPU E5-2697@2.70GHz and a 128.00GB Win7 internal memory 64-bit operating system, and the whole electrocardiosignal automatic classification algorithm is realized by adopting Matlab language.
In conjunction with fig. 1, 2 and 3, the implementation process of the present invention is as follows:
A) acquiring an electrocardiosignal of a human body, filtering, and detecting the R wave of the filtered electrocardiosignal:
firstly, signal acquisition: collecting an original electrocardiosignal of a human body at a collecting frequency of 1000Hz, storing the original electrocardiosignal into a data form of a TXT document, and reading the original electrocardiosignal data stored in the TXT document into a computer by Matlab software;
processing the electrocardio original signal data:
and finding out a QRS complex on a scale 4 through wavelet decomposition, and detecting the R wave peak point on the premise of determining the QRS complex. And (3) highlighting the QRS wave group again by adopting the transformation of an energy window, setting the size of the window to obtain different signal energy curves, searching for a peak value point of energy in the obtained energy curve, screening the predicted point of the R wave peak value obtained at the moment, and removing the unsatisfied condition so as to obtain the accurate position of the R wave peak value point.
Energy window transformation formula:
the formula of the satisfying condition of the R wave peak point is reserved: 0.4 RRmean<RR<1.6*RRmean。
B) And (3) similarity measurement:
and calculating the shortest distance between the selected data to be retrieved and the target data by adopting a TBD (tunnel boring device) improved DTW (data transfer threshold) method to represent the similarity between the two data, and marking the heartbeat data similar to the target data by setting a threshold delta.
C) Searching the electrocardiosignal characteristics of single heart beat:
firstly, heart beating is intercepted:
and taking 90 sampling points in front of the R wave position, taking 165 sampling points behind the R wave position to intercept the heartbeat data, and selecting the target heartbeat as a matching template.
And selecting the target heart beat as a matching template, selecting proper retrieval characteristics, and calculating the similarity of the feature vectors of the rest heart beat data and the target data.
Target data a and data b to be retrieved1,b2,...bmThe amplitude feature vector is itself; the morphological feature vector of the target data and the data to be retrieved represents the variation trend of each point by using the first derivative value of the electrocardiosignal at the point to form the morphological feature vector of the data.
And (3) calculating an expression of morphological feature vectors of the target data and the data to be retrieved:
② the experimental data selects 17 groups of data of No. 100, No. 102, No. 105, No. 106, No. 107, No. 109, No. 116, No. 119, No. 124, No. 209, No. 212, No. 214, No. 215, No. 220, No. 221, No. 232 and No. 234 in the MIT-BIH arrhythmia database.
And respectively adopting Euclidean distance, TBD, traditional DTW and the algorithm provided by the text to search amplitude and form so as to obtain the search positive detection rate and the sensitivity.
With reference to fig. 4 and 5, the implementation process of the present invention is as follows:
A) pretreatment:
B) and selecting proper retrieval characteristics, and calculating the similarity of the target heartbeat data and the heartbeat data to be retrieved.
Firstly, the electrocardiosignal retrieval with any length does not need to carry out heart shooting interception, the length of the electrocardiosignal retrieval can be selected at will, the data to be retrieved are divided according to the length of target data during data acquisition, and 20 sampling points are selected for retrieval through testing.
Secondly, intercepting the input electrocardiosignal data according to the length l of the target data a to obtain data b to be retrieved1,b2,...bmAnd searching the target data by calculating the similarity between the data to be searched and the target data. The two waveform characteristics of the amplitude characteristic and the morphological characteristic of the electrocardiosignal are still searched, and the whole experimental process is similar to that of single-heartbeat searching.
The electrocardiosignal retrieval algorithm can well retrieve target data, still has stronger stability compared with a common algorithm after noise is added, and the method is suitable for retrieval of electrocardiosignals.
Claims (4)
1. An electrocardiosignal retrieval method is characterized by comprising the following steps:
a. calculating the characteristics of the target data and the data to be retrieved;
b. improving the electrocardiosignal similarity measurement of the DTW based on the TBD;
c. evaluating the retrieval result by calculating the retrieval sensitivity and the positive retrieval rate;
d. retrieving amplitude characteristics of the electrocardiogram data;
e. retrieving morphological characteristics of the electrocardiogram data;
the specific process of the step a is as follows:
a-1, determining target data a with length l ═ a1,a2,...al]Wherein, the amplitude feature vector of a is itself;
a-2, intercepting data to be retrieved b on the electrocardiosignal X according to the length l of the target data a1,b2,...bmWhen intercepting the data to be retrieved, selecting the step length as k, and calculating the expression of the data to be retrieved as follows:
data b to be retrieved1,b2,...bmThe amplitude feature vector is itself;
a-3, the expression of the morphological feature vector of the target data is
k is a positive integer, and the expression of the morphological feature vector of the data to be retrieved is
k is a positive integer;
the specific process of the step b is as follows:
b-1, calculating the distance of the points on the two sequences by adopting TBD: selecting target data Q ═ Q1,q2,...qm) Calculating the data C ═ C (C) to be searched1,c2,...cn) Distance from each point of Q, TBD is defined as
Wherein the function f represents the corresponding electrocardiosignal waveform; then bring Q and C into
Obtaining a distance matrix D of m x n;
b-2, finding an optimal path from an S point to an E point in the distance matrix network, wherein the S point and the E point are two corresponding points on the target data and the data to be retrieved respectively; the cumulative distance r (i, j) for each point in the distance matrix D is the sum of itself and the shortest distance to this point, and the expression is:
r(i,j)＝dTBD(qi,cj)+min(r(i-1,j-1),r(i-1,j),r(i,j-1))；
b-3, calculating the shortest distance between the selected data to be retrieved and the target data by adopting a method of improving DTW (dynamic time warping) through TBD (tunnel boring machine) to represent the similarity between the two data, wherein the similarity definition formula of the improved algorithm is as follows
2. The method for retrieving electrocardiographic signals according to claim 1, wherein the specific process of step c is as follows: calculating the sensitivity and positive detection rate of the retrieval:
the sensitivity calculation formula is:
the positive detection rate is calculated by the formula:
wherein TP is the number of true positive samples, FN is the number of false negative samples, and FP is the number of false positive samples.
3. The method for retrieving an electrocardiographic signal according to claim 2, wherein in step d: the retrieval of the amplitude characteristics of the electrocardio data comprises the retrieval of the single-heart-beat electrocardio signal amplitude characteristics and the retrieval of the electrocardio signal amplitude characteristics with any length.
4. The method for retrieving an electrocardiographic signal according to claim 2, wherein in step e: the retrieval of the morphological characteristics of the electrocardio data comprises the retrieval of the morphological characteristics of the electrocardio signals of single heart beat and the retrieval of the morphological characteristics of the electrocardio signals with any length.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910611770.2A CN110169767B (en) | 2019-07-08 | 2019-07-08 | Retrieval method of electrocardiosignals |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910611770.2A CN110169767B (en) | 2019-07-08 | 2019-07-08 | Retrieval method of electrocardiosignals |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110169767A CN110169767A (en) | 2019-08-27 |
| CN110169767B true CN110169767B (en) | 2021-09-21 |
Family
ID=67699934
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910611770.2A Active CN110169767B (en) | 2019-07-08 | 2019-07-08 | Retrieval method of electrocardiosignals |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110169767B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111685759B (en) * | 2020-05-12 | 2022-09-09 | 河北大学 | P, T characteristic wave detection method of electrocardiosignals |
| CN115177267B (en) * | 2022-09-13 | 2023-01-03 | 合肥心之声健康科技有限公司 | Heart beat artifact identification method and system |
| CN116028914B (en) * | 2023-03-27 | 2023-06-16 | 深圳市魔样科技有限公司 | Intelligent finger ring identity authentication method and system |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5715829A (en) * | 1995-11-29 | 1998-02-10 | Hewlett-Packard Company | Method and apparatus for detecting heartbeats in an ECG waveform using an activity function and on-time search |
| CN1242693A (en) * | 1997-08-26 | 2000-01-26 | 精工爱普生株式会社 | Measuring, sensing and diagnosing apparatus and method relating to wave pulse, cardiac function, and motion intensity |
| CN103156599A (en) * | 2013-04-03 | 2013-06-19 | 河北大学 | Detection method of electrocardiosignal R characteristic waves |
| CN103345600A (en) * | 2013-06-24 | 2013-10-09 | 中国科学院深圳先进技术研究院 | Electrocardiosignal data processing method |
| CN106203324A (en) * | 2016-07-07 | 2016-12-07 | 中国矿业大学(北京) | The quick personal identification method of electrocardiosignal based on random tree |
| CN106889984A (en) * | 2017-01-22 | 2017-06-27 | 河北大学 | A kind of automatic noise-reduction method of electrocardiosignal |
| CN107390194A (en) * | 2017-07-20 | 2017-11-24 | 中国人民解放军国防科学技术大学 | A kind of radar target detection method based on the graceful divergence of full Donald Bragg |
| CN108883279A (en) * | 2016-04-06 | 2018-11-23 | 心脏起搏器股份公司 | The confidence level of arrhythmia detection |
| CN108926344A (en) * | 2018-07-26 | 2018-12-04 | 上海移视网络科技有限公司 | A kind of acute myocardial infarction AMI positioning automatic discrimination system based on CNN neural network |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7558622B2 (en) * | 2006-05-24 | 2009-07-07 | Bao Tran | Mesh network stroke monitoring appliance |
| CN104523266B (en) * | 2015-01-07 | 2017-04-05 | 河北大学 | A kind of electrocardiosignal automatic classification method |
| WO2017019184A2 (en) * | 2015-06-09 | 2017-02-02 | University Of Connecticut | Method and apparatus for removing motion artifacts from biomedical signals |
-
2019
- 2019-07-08 CN CN201910611770.2A patent/CN110169767B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5715829A (en) * | 1995-11-29 | 1998-02-10 | Hewlett-Packard Company | Method and apparatus for detecting heartbeats in an ECG waveform using an activity function and on-time search |
| CN1242693A (en) * | 1997-08-26 | 2000-01-26 | 精工爱普生株式会社 | Measuring, sensing and diagnosing apparatus and method relating to wave pulse, cardiac function, and motion intensity |
| CN103156599A (en) * | 2013-04-03 | 2013-06-19 | 河北大学 | Detection method of electrocardiosignal R characteristic waves |
| CN103345600A (en) * | 2013-06-24 | 2013-10-09 | 中国科学院深圳先进技术研究院 | Electrocardiosignal data processing method |
| CN108883279A (en) * | 2016-04-06 | 2018-11-23 | 心脏起搏器股份公司 | The confidence level of arrhythmia detection |
| CN106203324A (en) * | 2016-07-07 | 2016-12-07 | 中国矿业大学(北京) | The quick personal identification method of electrocardiosignal based on random tree |
| CN106889984A (en) * | 2017-01-22 | 2017-06-27 | 河北大学 | A kind of automatic noise-reduction method of electrocardiosignal |
| CN107390194A (en) * | 2017-07-20 | 2017-11-24 | 中国人民解放军国防科学技术大学 | A kind of radar target detection method based on the graceful divergence of full Donald Bragg |
| CN108926344A (en) * | 2018-07-26 | 2018-12-04 | 上海移视网络科技有限公司 | A kind of acute myocardial infarction AMI positioning automatic discrimination system based on CNN neural network |
Non-Patent Citations (3)
| Title |
|---|
| Dynamic time warp distances as feedback for EEG feature density;Ward, C.R.;《2015 IEEE Signal Processing in Medicine and Biology Symposium (SPMB)》;20151231;第1-5页 * |
| Total Bregman Divergence and its Applications to Shape Retrieval;Liu, Meizhu;《Total Bregman Divergence and its Applications to Shape Retrieval》;20101231;第3463-3468页 * |
| 基于PLR-DTW的ECG身份识别方法;杨立才;《生物医学工程学杂志》;20131025;第976-981页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110169767A (en) | 2019-08-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110169767B (en) | Retrieval method of electrocardiosignals | |
| de Albuquerque et al. | Robust automated cardiac arrhythmia detection in ECG beat signals | |
| CN108511055B (en) | Ventricular premature beat recognition system and method based on classifier fusion and diagnosis rules | |
| Alaskar et al. | The implementation of pretrained AlexNet on PCG classification | |
| Lu et al. | QRS detection based on improved adaptive threshold | |
| Kaya et al. | Feature selection using genetic algorithms for premature ventricular contraction classification | |
| Bono et al. | Development of an automated updated selvester QRS scoring system using SWT-based QRS fractionation detection and classification | |
| WO2019161611A1 (en) | Ecg information processing method and ecg workstation | |
| Chatterjee et al. | Real time P and T wave detection from ECG using FPGA | |
| Saini et al. | Delineation of ecg wave components using k-nearest neighbor (knn) algorithm: Ecg wave delineation using knn | |
| Zifan et al. | Automated ECG segmentation using piecewise derivative dynamic time warping | |
| CN112932498B (en) | T waveform state classification system with generalization capability based on deep learning | |
| Wang et al. | Automatic digital ECG signal extraction and normal QRS recognition from real scene ECG images | |
| Zhang et al. | An ECG heartbeat classification method based on deep convolutional neural network | |
| CN112107310A (en) | ECG identity recognition method based on IWT and AGA-BP models | |
| CN111180071B (en) | Method and device for calculating relationship between high-risk HPV type and cervical cancer pre-lesion stage | |
| CN111053552B (en) | QRS wave detection method based on deep learning | |
| Mei et al. | YOLO-lung: A practical detector based on imporved YOLOv4 for Pulmonary Nodule Detection | |
| Gao et al. | Real time ECG characteristic point detection with randomly selected signal pair difference (RSSPD) feature and random forest classifier | |
| Sehirli et al. | A novel method for segmentation of QRS complex on ECG signals and classify cardiovascular diseases via a hybrid model based on machine learning | |
| Liu et al. | A novel P-QRS-T wave localization method in ECG signals based on hybrid neural networks | |
| CN113069124A (en) | Electrocardio monitoring method based on CNN-ET model | |
| Wu et al. | Optimized pan-tompkins based heartbeat detection algorithms | |
| Krimi et al. | An approach combining wavelet transform and hidden Markov models for ECG segmentation | |
| CN116521912B (en) | Ultrasonic data storage management system and method based on artificial intelligence |
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 |