CN103948387A - Polymorphism reconstruction and optimization method for realizing abnormal electrocardiogram template based on big data - Google Patents
Polymorphism reconstruction and optimization method for realizing abnormal electrocardiogram template based on big data Download PDFInfo
- Publication number
- CN103948387A CN103948387A CN201410097341.5A CN201410097341A CN103948387A CN 103948387 A CN103948387 A CN 103948387A CN 201410097341 A CN201410097341 A CN 201410097341A CN 103948387 A CN103948387 A CN 103948387A
- Authority
- CN
- China
- Prior art keywords
- template
- kinds
- waveform
- wave form
- ripples
- 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.)
- Granted
Links
Abstract
The invention provides a polymorphism reconstruction and optimization method for realizing an abnormal electrocardiogram template based on big data. The polymorphism reconstruction and optimization method is characterized by comprising the following steps that firstly, parameter template data with two-dimension characteristics are built according to decomposing parameters of standard electrocardiogram waveforms, and an electrocardiogram waveform template consisting of a plurality of fine nodes is built; then, the big data of the existing and edited dynamic electrocardiogram monitor are utilized for preprocessing before comparison, i.e., waveform segmentation, and next, the gradual waveform comparison with the electrocardiogram waveform template is carried out; next, three key quantities including FRR (fault read rate), FAR (false accept rate) and TH of the value range being 0 to 1 are used for counting the success rate finally matched with the assessment in a comparison algorithm, in addition, the polymorphism reconstruction is carried out, or the template data volume optimization is carried out again.
Description
Technical field
What the present invention relates to is a kind of polymorphic reconstruct and automatic optimization method of realizing abnormal electrocardiogram template based on large data, be mainly to use the large data of dynamic electro-cardiac monitor generation to carry out the method for dynamic construction, optimization abnormal electrocardiogram data template variform, belong to dynamic electrocardiogram monitoring technical field.
Background technology
At present, in dynamic electrocardiogram monitoring in medical treatment, identification based on abnormal electrocardiographic pattern is carried out manual evaluation by medical practitioner often, not only efficiency is low but also hand labor amount is large for it, although some medical special electrocardio software or equipment also can be realized dynamic pre-identification, but misclassification rate and to refuse to recognize rate higher, its reason is that most of electrocardio recognizers are the abnormal electrocardiogram data templates based on conventional criteria, differ greatly for different individual humans, such as individual variations such as old man, child, youngster, middle age, the display form of abnormal electrocardiogram can be different.Therefore for dynamic electrocardiogram monitoring, each action of human body can present more significantly diversity, and in this case, obviously standard cardioelectric template is the actual needs that can not meet dynamic electrocardiogram monitoring.
Summary of the invention
The object of the invention is to overcome the deficiency that prior art exists, and provide a kind of be mainly the large data of using dynamic electro-cardiac monitor to generate carry out dynamic construction, optimize abnormal electrocardiogram data template variform realize polymorphic reconstruct and the optimization method of abnormal electrocardiogram template based on large data.
The object of the invention is to complete by following technical solution, described polymorphic reconstruct and the optimization method of realizing abnormal electrocardiogram template based on large data, it comprises the steps:
First, according to the resolution parameter of standard cardioelectric waveform, set up the parameterized template data with two-dimensional characteristics, and establish the ecg wave form template being formed by multiple minutiae point;
Secondly, utilize large data existing and edited dynamic electro-cardiac monitor, the pretreatment before comparing, carries out waveform merogenesis, then carries out waveform one by one with ecg wave form template and compares;
Again, on alignment algorithm, use misclassification rate (FRR), refuse to recognize rate (FAR), span: these three critical quantity of 0～1 similarity (TH) are added up and assessed final success rate of mating, and make polymorphic reconstruct or re-start the optimization of template data amount.
Described ecg wave form template of having established comprises: 2 kinds of cardiac electric axis, 2 kinds of PR intervals, 2 kinds of QTc intervals, 3 kinds of R ripples, 3 kinds of S ripples, 5 kinds of P ripples, 5 kinds of T ripples, 12 kinds of Q ripples, 36 kinds of ST ripples, 30 kinds of ST-T associatings.
Comparing between described ecg wave form template and arbitrarily EGC waveform data, by by misclassification rate, default three groups of threshold values of refusing to recognize rate, these three critical quantity of similarity carry out three times and calculate, and according to three statistical result, assess as follows:
(1) according to the similarity TH contrast in th result and last group calculated for the third time, for the ecg wave form lower than threshold value, be reconstructed;
(2) contrast with the similarity TH in second group according to the th result of calculating for the second time, for identifying but the very low template of discrimination, the optimization of again entering new template data volume;
(3) contrast with the similarity TH in first group according to the th result of calculating for the first time, for identifying but the very low template of discrimination, the optimization of again entering new template data volume; For the high waveform of discrimination, template is not carried out any processing.
The present invention is by the large data scanning to existing cardioelectric monitor, and EGC pattern variant class template can get more and more, classification is more and more thinner, and waveform template is also more and more accurate; Improve the quality of waveform template, nature can improve the accuracy of the structure of electrocardio disease template, and this accuracy rate for dynamic electrocardiogram monitoring can improve a lot.
Brief description of the drawings
Fig. 1 is ecg wave form template configuration classification schematic diagram of the present invention.
Fig. 2 is ecg wave form template of the present invention reconstruct and optimization schematic diagram.
Detailed description of the invention
Below in conjunction with drawings and Examples, the present invention will be further described: polymorphic reconstruct and the optimization method of realizing abnormal electrocardiogram template based on large data of the present invention, it comprises the steps:
First, according to the resolution parameter of standard cardioelectric waveform, set up the parameterized template data with two-dimensional characteristics, and establish the ecg wave form template being formed by multiple minutiae point;
Secondly, utilize large data existing and edited dynamic electro-cardiac monitor, the pretreatment before comparing, carries out waveform merogenesis, then carries out waveform one by one with ecg wave form template and compares;
Again, on alignment algorithm, use misclassification rate (FRR), refuse to recognize rate (FAR), span: these three critical quantity of 0～1 similarity (TH) are added up and assessed final success rate of mating, and make polymorphic reconstruct or re-start the optimization of template data amount.
The ecg wave form template of described establishment comprises: 2 kinds of cardiac electric axis, 2 kinds of PR intervals, 2 kinds of QTc intervals, 3 kinds of R ripples, 3 kinds of S ripples, 5 kinds of P ripples, 5 kinds of T ripples, 12 kinds of Q ripples, 36 kinds of ST ripples, 30 kinds of ST-T associatings.
Comparing between described ecg wave form template and arbitrarily EGC waveform data, by by misclassification rate, default three groups of threshold values of refusing to recognize rate, these three critical quantity of similarity carry out three times and calculate, and according to three statistical result, assess as follows:
(1) according to the similarity TH contrast in th result and last group calculated for the third time, for the ecg wave form lower than threshold value, be reconstructed;
(2) contrast with the similarity TH in second group according to the th result of calculating for the second time, for identifying but the very low template of discrimination, the optimization of again entering new template data volume;
(3) contrast with the similarity TH in first group according to the th result of calculating for the first time, for identifying but the very low template of discrimination, the optimization of again entering new template data volume; For the high waveform of discrimination, template is not carried out any processing.
Embodiment: polymorphic reconstruct and the optimization method of realizing abnormal electrocardiogram template based on large data of the present invention, it relates to structure electrocardio medical diagnosis on disease and two kinds of data templates of ecg wave form; Wherein electrocardio medical diagnosis on disease template is to be made up of one group of multiple ecg wave form template that form a kind of genius morbi that have that are associated; Described ecg wave form template is made up of multiple minutiae point; The object of utilizing large data is that reconstruct is new or optimize existing ecg wave form template data, as shown in Figure 1.
First according to the resolution parameter of standard cardioelectric waveform, set up the parameterized template data with two-dimensional characteristics, established at present following ecg wave form template: cardiac electric axis (2 kinds), PR interval (2 kinds), QTc interval (2 kinds), R ripple (3 kinds), S ripple (3 kinds), P ripple (5 kinds), T ripple (5 kinds), Q ripple (12 kinds), ST ripple (36 kinds), ST-T combine (30 kinds); Observation process by the heart patient to a large amount of sees, the template kind of waveform is not enough far away, therefore need constantly find and add for a long time, and for existing electrocardio template, more sophisticated category.
According to the existing and large data of the dynamic electro-cardiac monitor of edited (by effective waveform montage in 5 minutes), the pretreatment before first comparing, carries out waveform merogenesis, then carries out waveform one by one with ecg wave form template and compares.On alignment algorithm we use misclassification rate (FRR), refuse to recognize rate (FAR), similarity (TH) (span: 0～1) three critical quantity are added up and assess final success rate of mating.Arbitrarily comparing between EGC waveform data and waveform template, by by misclassification rate, three groups of default threshold values of refusing to recognize rate, three critical quantity of similarity carry out three calculating, according to three statistical result, assess as follows:
(1) according to the similarity TH contrast in th result and last group calculated for the third time, for the ecg wave form lower than threshold value, be reconstructed;
(2) contrast with the similarity TH in second group according to the th result of calculating for the second time, for identifying but the very low template of discrimination, the optimization of again entering new template data volume;
(3) contrast with the similarity TH in first group according to the th result of calculating for the first time, for identifying but the very low template of discrimination, the optimization of again entering new template data volume.For the high waveform of discrimination, template is not carried out any processing.
The monitoring of dynamic electrocardiogram in Fig. 1, diagnose template realization by various diseases, each medical diagnosis on disease template all have oneself 4 No. ID.What for example, we determined acute myocardial infarction is for No. ID 8012.In each medical diagnosis on disease template, comprise one group of ecg wave form template, each ecg wave form template also forms for No. ID by 6, and front 2 is type code, and latter 4 is ecg wave form template code.For example the ecg wave form template data in acute myocardial infarction is expressed as:
8012：T0003；
Q0501；
R0341；
PR0005；
ST2232；
……
Ecg wave form template in Fig. 1, is divided equidistant piece with time shaft and is formed by multiple, and each piece is made up of one group of minutiae point, and each minutiae point is made up of a stack features value; Equidistant piece, minutiae point and eigenvalue are by 6 distinguished for No. ID, and explanatory content is the same.
The foundation of electrocardio medical diagnosis on disease template needs expert doctor to complete, cannot automatically be reconstructed by computer, therefore the present invention mainly refers to the structure of new waveform template and assessment and the optimization of existing waveform template data that ecg wave form template is carried out under the large data environment of dynamic ecg monitoring, so following example, taking ecg wave form template as objective for implementation.
1) set up the data structure of ecg wave form template
2) to Electrocardiographic waveform pre-treatment step:
● baseline drift processing
● cardiac electrical cycle is divided second
● waveform is cut apart
● piece is divided
●…...
3) realization of alignment algorithm:
First we need to determine that the predetermined amount of refusing to recognize rate (FAR), misclassification rate (FRR) and similarity (Similarity) three coupling is three groups:
0.1%、0.02%、0.75%;
0.05%、0.06%、0.85%;
0.03%、0.08%、0.93%;
Also need, to the every important indicator in waveform, piece, minutiae point, to set up rational comparison threshold value by a large amount of measuring and calculating, do not specify at this simultaneously.
The ECGWaveVerify (WaveData, WaveTemplate, VerifyParameter) of alignment algorithm;
Alignment algorithm flow process is shown in that Fig. 2 is as shown:
Fig. 2 is ecg wave form template of the present invention reconstruct and Optimizing Flow figure;
4) ecg wave form template reconstruct:
Template reconstruction of function is by name:
ECGWaveTemplateRefactoring(OldWaveTemplate,NewWaveTemplate,Parameter);
5) ecg wave form is template optimized:
Template optimized function is by name:
ECGWaveTemplateOptimize(WaveTemplate,Parameter);
Effect of the present invention just designs from the angle of electrocardio template, can improve to a certain extent precision and the reliability of ecg wave form template, promotes the accuracy rate of dynamic electrocardiogram monitoring.
Claims (3)
1. realize polymorphic reconstruct and the optimization method of abnormal electrocardiogram template based on large data, it is characterized in that described polymorphic reconstruct and optimization method are:
First, according to the resolution parameter of standard cardioelectric waveform, set up the parameterized template data with two-dimensional characteristics, and establish the ecg wave form template being formed by multiple minutiae point;
Secondly, utilize large data existing and edited dynamic electro-cardiac monitor, the pretreatment before comparing, carries out waveform merogenesis, then carries out waveform one by one with ecg wave form template and compares;
Again, on alignment algorithm, use misclassification rate (FRR), refuse to recognize rate (FAR), span: these three critical quantity of 0～1 similarity (TH) are added up and assessed final success rate of mating, and make polymorphic reconstruct or re-start the optimization of template data amount.
2. polymorphic reconstruct and the optimization method of realizing abnormal electrocardiogram template based on large data according to claim 1, the ecg wave form template of having established described in it is characterized in that comprises: 2 kinds of cardiac electric axis, 2 kinds of PR intervals, 2 kinds of QTc intervals, 3 kinds of R ripples, 3 kinds of S ripples, 5 kinds of P ripples, 5 kinds of T ripples, 12 kinds of Q ripples, 36 kinds of ST ripples, 30 kinds of ST-T associatings.
3. polymorphic reconstruct and the optimization method of realizing abnormal electrocardiogram template based on large data according to claim 1, it is characterized in that comparing between described ecg wave form template and any EGC waveform data, by by misclassification rate, default three groups of threshold values of refusing to recognize rate, these three critical quantity of similarity carry out three times and calculate, and according to three statistical result, assess as follows:
(1) according to the similarity TH contrast in th result and last group calculated for the third time, for the ecg wave form lower than threshold value, be reconstructed;
(2) contrast with the similarity TH in second group according to the th result of calculating for the second time, for identifying but the very low template of discrimination, the optimization of again entering new template data volume;
(3) contrast with the similarity TH in first group according to the th result of calculating for the first time, for identifying but the very low template of discrimination, the optimization of again entering new template data volume; For the high waveform of discrimination, template is not carried out any processing.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410097341.5A CN103948387B (en) | 2014-03-17 | 2014-03-17 | A kind of polymorphic reconstruct and optimization method realizing abnormal electrocardiogram template based on large data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410097341.5A CN103948387B (en) | 2014-03-17 | 2014-03-17 | A kind of polymorphic reconstruct and optimization method realizing abnormal electrocardiogram template based on large data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103948387A true CN103948387A (en) | 2014-07-30 |
| CN103948387B CN103948387B (en) | 2016-01-20 |
Family
ID=51325847
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410097341.5A Active CN103948387B (en) | 2014-03-17 | 2014-03-17 | A kind of polymorphic reconstruct and optimization method realizing abnormal electrocardiogram template based on large data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103948387B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104473633A (en) * | 2014-12-31 | 2015-04-01 | 广州视源电子科技股份有限公司 | Judging method and device of abnormal electrocardio data |
| CN110731762A (en) * | 2019-09-18 | 2020-01-31 | 平安科技（深圳）有限公司 | Method, device, computer system and readable storage medium for preprocessing pulse wave based on similarity |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1189320A (en) * | 1997-01-31 | 1998-08-05 | 惠普公司 | Method and system for fast determination of EKG waveform morphology |
| US7177674B2 (en) * | 2001-10-12 | 2007-02-13 | Javier Echauz | Patient-specific parameter selection for neurological event detection |
| US20100152598A1 (en) * | 2008-12-11 | 2010-06-17 | Siemens Medical Solutions Usa, Inc. | System for Heart Performance Characterization and Abnormality Detection |
| CN101810476A (en) * | 2009-12-22 | 2010-08-25 | 李顶立 | Classification method of heart beat template of dynamic electrocardiogram |
-
2014
- 2014-03-17 CN CN201410097341.5A patent/CN103948387B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1189320A (en) * | 1997-01-31 | 1998-08-05 | 惠普公司 | Method and system for fast determination of EKG waveform morphology |
| US7177674B2 (en) * | 2001-10-12 | 2007-02-13 | Javier Echauz | Patient-specific parameter selection for neurological event detection |
| US20100152598A1 (en) * | 2008-12-11 | 2010-06-17 | Siemens Medical Solutions Usa, Inc. | System for Heart Performance Characterization and Abnormality Detection |
| CN101810476A (en) * | 2009-12-22 | 2010-08-25 | 李顶立 | Classification method of heart beat template of dynamic electrocardiogram |
Non-Patent Citations (2)
| Title |
|---|
| 曹国超: "动态心电图波形改进分类策略研究", 《中国优秀硕士学位论文全文数据库信息科技辑》, 31 July 2010 (2010-07-31) * |
| 杨洁: "心电算法及其参数数据库的构建", 《中国优秀硕士学位论文全文数据库医药卫生科技辑》, 31 January 2005 (2005-01-31) * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104473633A (en) * | 2014-12-31 | 2015-04-01 | 广州视源电子科技股份有限公司 | Judging method and device of abnormal electrocardio data |
| CN110731762A (en) * | 2019-09-18 | 2020-01-31 | 平安科技（深圳）有限公司 | Method, device, computer system and readable storage medium for preprocessing pulse wave based on similarity |
| CN110731762B (en) * | 2019-09-18 | 2022-02-08 | 平安科技（深圳）有限公司 | Method, device, computer system and readable storage medium for preprocessing pulse wave based on similarity |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103948387B (en) | 2016-01-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Gao et al. | An effective LSTM recurrent network to detect arrhythmia on imbalanced ECG dataset | |
| Çınar et al. | Classification of normal sinus rhythm, abnormal arrhythmia and congestive heart failure ECG signals using LSTM and hybrid CNN-SVM deep neural networks | |
| CN109171712B (en) | Atrial fibrillation identification method, atrial fibrillation identification device, atrial fibrillation identification equipment and computer readable storage medium | |
| CN110443798B (en) | Autism detection method, device and system based on magnetic resonance image | |
| CN110090012A (en) | A kind of human body diseases detection method and testing product based on machine learning | |
| US20230143594A1 (en) | Systems and methods for reduced lead electrocardiogram diagnosis using deep neural networks and rule-based systems | |
| CN103714281A (en) | Identity recognition method based on electrocardiosignals | |
| Kaur et al. | Feature extraction and principal component analysis for lung cancer detection in CT scan images | |
| Zhang et al. | A multi-dimensional association information analysis approach to automated detection and localization of myocardial infarction | |
| CN104840186A (en) | Evaluation method of autonomic nervous function of patient suffering from CHF (Congestive Heart-Failure) | |
| Yan et al. | A hybrid outlier detection method for health care big data | |
| Manju et al. | Classification of cardiac arrhythmia of 12 lead ecg using combination of smoteenn, xgboost and machine learning algorithms | |
| EP4041073A1 (en) | Systems and methods for electrocardiogram diagnosis using deep neural networks and rule-based systems | |
| Mjahad et al. | Detection of ventricular fibrillation using the image from time-frequency representation and combined classifiers without feature extraction | |
| Yang et al. | A novel method for automated congestive heart failure and coronary artery disease recognition using THC-Net | |
| Zhang et al. | Application of deep neural network for congestive heart failure detection using ECG signals | |
| CN103948387A (en) | Polymorphism reconstruction and optimization method for realizing abnormal electrocardiogram template based on big data | |
| CN112690802B (en) | Method, device, terminal and storage medium for detecting electrocardiosignals | |
| Kelarev et al. | Improving Classifications for Cardiac Autonomic Neuropathy Using Multi-level Ensemble Classifier and Feature Selection Based on Random Forest | |
| Chen et al. | Cross-modal multiscale multi-instance learning for long-term ECG classification | |
| CN108338777A (en) | A kind of pulse signal determination method and device | |
| Deng et al. | Classification of cardiac electrical signals between patients with myocardial infarction and normal subjects by using nonlinear dynamics features and different classification models | |
| Xie et al. | Discrete Biorthogonal Wavelet Transform Based Convolutional Neural Network for Atrial Fibrillation Diagnosis from Electrocardiogram. | |
| Huang et al. | A congestive heart failure detection system via multi-input deep learning networks | |
| CN115015683B (en) | Cable production performance test method, device, equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |