CN101461710B - Shockable rhythm recognition algorithm based on grid projection distribution dispersion - Google Patents
Shockable rhythm recognition algorithm based on grid projection distribution dispersion Download PDFInfo
- Publication number
- CN101461710B CN101461710B CN2009100451537A CN200910045153A CN101461710B CN 101461710 B CN101461710 B CN 101461710B CN 2009100451537 A CN2009100451537 A CN 2009100451537A CN 200910045153 A CN200910045153 A CN 200910045153A CN 101461710 B CN101461710 B CN 101461710B
- Authority
- CN
- China
- Prior art keywords
- grid projection
- heart
- distribution dispersion
- rhythm
- grid
- 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.)
- Expired - Fee Related
Links
Images
Landscapes
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
- Electrotherapy Devices (AREA)
Abstract
A recognition algorithm of heart rhythm capable of electric shock cardioversion based on grid projection distribution dispersion, is suitable for disease diagnosis and treatment instrument or apparatus, including the steps: S1, preprocessing the electrocardiosignal; S2, recognizing whether the electrocardiosignal is the heart rhythm of cardiac arrest, if so, judging the heart rhythm as the heart rhythm incapable of electric shock cardioversion, otherwise, continuing the follow-up steps S3 and S4; S3, calculating the distribution dispersion of the grid projection; S4, determining whether it pertains to the heart rhythm incapable of electric shock cardioversion or the heart rhythm capable of electric shock cardioversion based on the distribution dispersion of the grid projection. The invention improves the sensitivity and specificity of recognizing the heart rhythm capable of electric shock cardioversion, simplifies the computational complexity of the algorithm, can be applied to the existing ECG guardianship equipment and automatic external defibrillator, and the like instruments requiring body surface electrocardiogram for recognizing the heart rhythm capable of electric shock cardioversion.
Description
Technical field
The present invention relates to a kind of electrocardiosignal (ECG) identification instrument, but particularly a kind of improved the Electrical Cardioversion rhythm of the heart (Shockable Rhythm, ShR) identification instrument that is used for electrocardiogram monitor and automated external defibrillator.
Background technology
Sudden cardiac death (SCD) is meant the natural death of the unexpected generation that causes owing to the heart reason.The reason major part that causes sudden cardiac death is momentary dysfunction and the electrophysiological change that takes place on all kinds of cardiovascular pathological changes basis, and cause that malignant ventricular arrhythmia such as ventricular tachycardia (are called for short chamber speed, VT), ventricular fibrillation (is called for short the chamber and quivers, VF) etc.Electric defibrillation is the first-selected effective ways that stop most rapidity malignant ventricular arrhythmias.
1997, American Heart Association (AHA) has delivered a suggestion relevant with automated external defibrillator (AED) algorithm performance report " automated external defibrillator that is used for the public arena defibrillation: to the performance of explanation and report arrhythmia analysis algorithm on circulation (Circulation) magazine, comprise the suggestion of new waveform and raising safety " (" Automatic External Defibrillators forPublic Access Defibrillation:Recommendations for Specifying and ReportingArrhythmia Analysis Algorithm Performance, Incorporating New Waveforms, and Enhancing Safety. ").
This suggestion is divided into following three major types with the rhythm of the heart: but the Electrical Cardioversion rhythm of the heart (shockable rhythms, ShR), can not the Electrical Cardioversion rhythm of the heart (nonshockable rhythms, NShR) and the middle rhythm of the heart (Intermediate rhythms).
At present but the Electrical Cardioversion rhythm recognition algorithm of bibliographical information exists variety of issue, as since the chamber when quivering Electrocardiographic form can change a lot, but various algorithm based on ECG R wave identification is not suitable for the differentiation of the Electrical Cardioversion rhythm of the heart; Phase space rebuild (Phase Space ReconstructionAlgorithm, PSR) algorithm, signal comparison algorithm (Signal Comparison Algorithm, though SCA) wait very high specificity is arranged, sensitivity is very poor; And some are based on the algorithm computation complexity of various conversion and analysis of complexity, to having relatively high expectations of hardware.So, but the differentiation algorithm of the existing Electrical Cardioversion rhythm of the heart still exists sensitivity and specificity not to take into account, or problem such as calculation of complex, for example, as typical example, also there are some such shortcomings in the HILB algorithm application in the instrument of the diagnosis and treatment of disease or device, the HILB algorithm has used method-Hilbert transform method of often using when analyzing nonlinear properties to make up phase space.Suppose that electrocardiosignal is x (t), obtain x after it is done Hilbert transform
H(t), if, use x with x (t) expression x axial coordinate
H(t) represent the y axial coordinate, just constructed the phase space of a two dimension.In such phase space, the track of chaotic signal can be more mixed and disorderly than the track of rule signal.People such as Anoton, Robert and Karl find that the trajectory of phase space of VF signal is more mixed and disorderly than the trajectory of phase space of SR (sinus rhythm) signal.So they suppose that the VF signal is a chaos, and the SR signal is a rule.They are divided into the grid of 40x40 identical size with the phase space that builds, and the grid of the trajectory of phase space process of statistics electrocardiosignal is counted.Because the SR signal is a rule, the VF signal is a chaos, so compare with the trajectory of phase space of SR signal, the trajectory of phase space of VF signal can pass through more grid.
In order to reduce amount of calculation, also need signal is done down-sampled.
The detailed process of HILB algorithm is as follows:
1. down-sampled with 50Hz to signal.
2. the Hilbert transform of electrocardiosignal x (t) is x
H(t), make up the phase space of 40x40 lattice, calculate (x (t), x
H(t)) shared lattice are counted visited boxes in constructed phase space.
3. definition
And to get threshold value be d0,
If d＞d0 then is judged to VF;
If d＜=d0 then is judged to SR.
Summary of the invention:
As mentioned above, but for electrocardiogram monitor and automated external defibrillator provide the Electrical Cardioversion rhythm recognition algorithm of discriminant accuracy height and fast operation, be technical problem to be solved by this invention.For this reason, but the object of the present invention is to provide a kind of discern accurately, calculate simple, can satisfy application requirements, based on the Electrical Cardioversion rhythm of the heart identification instrument of grid projection distribution dispersion, but to improve the existing performance that needs to use the instrument and equipment of Electrical Cardioversion rhythm of the heart recognition methods.
Technical scheme of the present invention is as follows:
But, comprise as follows according to a kind of improved Electrical Cardioversion rhythm of the heart identification instrument of the present invention based on grid projection distribution dispersion:
At first, electrocardiosignal is carried out the identification of the asystole rhythm of the heart:
If the asystole rhythm of the heart then is judged to NShR;
If not the asystole rhythm of the heart, then carry out the step of back.
Secondly, calculate grid projection distribution dispersion;
At last, differentiate NShR and ShR according to grid projection distribution dispersion,
Discrimination standard is:
If grid projection distribution dispersion＞=threshold value then is judged to NShR;
If grid projection distribution dispersion＜threshold value then is judged to ShR.
The detailed process of the above-mentioned identification asystole rhythm of the heart is:
Amplitude is judged to the asystole rhythm of the heart less than the electrocardiosignal of 80uV.
The detailed process of aforementioned calculation grid projection distribution dispersion is:
At first, one section electrocardiogram (ECG) data is divided into segment by identical interval, each segment is called a grizzly bar (bar), and each interval is called grill width (barwidth);
Then, the amplitude range that calculating ECG covers in each grizzly bar is exactly the projection (shadow) on grizzly bar y axle;
Then, between the maximum and minima in all grid projections (shadow), divide several (histnum) zones, statistics drops on the quantity (shadow hist) of the grid projection in each zone;
At last, calculate the standard deviation (shadow stdhist) of this histnum shadow hist, the dispersion of grid projection distribution just.
Owing to adopted above technical scheme, but improved the sensitivity and the specificity of the identification Electrical Cardioversion rhythm of the heart, satisfied application requirements.Also simplified the computation complexity of algorithm in addition.The present invention can be applicable to electrocardiogram monitor and automated external defibrillator (AED) but etc. need be according to the instrument and equipment of the surface electrocardiogram identification Electrical Cardioversion rhythm of the heart.
Description of drawings:
Fig. 1 is agent structure of the present invention and flow chart.
Fig. 2 is the flow chart of " S1 pretreatment " module among the main process figure of the present invention.
Fig. 3 is the flow chart of " S3 calculates grid projection distribution dispersion " module among the main process figure of the present invention.
The specific embodiment:
The invention will be further described below by specific embodiment.
Present embodiment is that the present invention is at personal computer (PC) and matrix experiment chamber (MatrixLaboratory, Matlab) a kind of possible realization on the platform, and on the test data set that constitutes by three standard databases of the arrhythmia data base of Massachusetts Polytechnics (MITDB), the ventricular arrhythmia data base of Ke Laideng university (CUDB), the malignant ventricular arrhythmia data base of Massachusetts Polytechnics (VFDB), test and compare.The present embodiment concrete steps are as follows:
1. electrocardiosignal is carried out pretreatment:
A) moving average filter on one 5 rank of use, high-frequency noises such as filtering spread noise and myoelectricity noise;
B) use the high pass filter of a cut-off frequency, suppress baseline drift as 1Hz;
C) use the Butterworth low pass filter of a cut-off frequency, further the irrelevant radio-frequency component of filtering as 30Hz.
2. electrocardiosignal is carried out the identification of the asystole rhythm of the heart:
If the amplitude of electrocardiosignal less than 80uV, is then thought the asystole rhythm of the heart, be judged to NShR;
Not the asystole rhythm of the heart if the amplitude of electrocardiosignal more than or equal to 80uV, is then thought, continue the step of back.
3. calculate grid projection distribution dispersion:
A) one section electrocardiogram (ECG) data is divided into segment by identical interval, each segment is called a grizzly bar (bar), and interval is called grill width (barwidth).Barwidth is taken as 24ms (when sample rate is 250Hz, corresponding to 6 sampled points);
B) amplitude range that covers in each grizzly bar of calculating ECG is exactly the projection (shadow) on grizzly bar y axle;
C) between the maximum and minima in all grid projections (shadow), divide several (histnum) zones, statistics drops on the quantity (shadow_hist) of the grid projection in each zone.Wherein, histnum is taken as 10;
D) calculate the standard deviation (shadow_stdhist) of this histnum shadow_hist, the dispersion of grid projection distribution just.
4. differentiate NShR and ShR according to standardization grid projection standard deviation:
Discrimination standard is:
If standardization grid projection standard deviation=threshold value T then is judged to NShR;
If standardization grid projection standard deviation＜threshold value T then is judged to ShR.
The software and hardware configuration that present embodiment uses is as follows:
-hardware: Dell is to 4 computers, dominant frequency 226GHz, 512,000,000 internal memories (Dell OPTIPLEXGX270, Pentium (R) 4 (2.26GHz) and 512MB DDR SDRAM)
-software: MATLAB R13, " signal processing workbox " version 6.0 (" Signal ProcessingToolbox " version 6.0)
Under following test condition, to present embodiment and prior art Hilbert (HILB) algorithm
[1] [2]Test and compare:
Test data set is all data of MITDB, CUDB, three standard databases of VFDB, is a segment (sample data) with 8s, and adjacent two segment zero-times differ 1s.
The goldstandard (Golden Standard) of rhythm of the heart classification:
A) the reference note that carries according to the data base (reference annotation) carries out rhythm of the heart classification to the data segment.
B) ShR: the rhythm of the heart (rhythm) class annotation information is labeled as the electrocardiogram (ECG) data of VF, VT,
NShR: other all rhythms of the heart;
C) containing the segment of mixing the rhythm of the heart does not use.
Test result such as following table:
Wherein, AUC is meant and receives operating characteristic (ROC) area under a curve
[3] [4], be concentrated expression sensitivity and specific index.
By in the table as seen, the AUC of present embodiment (0.980) is greater than the AUC (0.965) of HILB algorithm, and remarkable on this difference statistical significance
。The classification performance that present embodiment is described is better than the HILB algorithm.And also be less than the HILB algorithm computation time of present embodiment.
If threshold value T is taken as 58.1, but in the present embodiment based on the sensitivity of the Electrical Cardioversion rhythm recognition algorithm of grid projection distribution dispersion be 92.1%, specificity is 95%, reaches the sensitivity 90% that AHA advises, the performance requirement of specificity 95%.
* list of references of the present invention
[1]DI?Robert?Tratnig.Reliability?of?New?Fibrillation?DetectionAlgorithms?for?Automated?External?Defibrillators[D].Dornbirn，Austria：Technische?Universit¨at?Graz，2005.
[2]A.Amann，R.Tratnig，K.Unterkofler.A?new?ventricular?fibrillationdetection?algorithm?for?automated?external?defibrillators[J].Computers?inCardiology，2005：559-562.
[3] JP Marques work, Wu Yifei translates. pattern recognition---principle, method and application [M]. and publishing house of Tsing-Hua University, 2002:113-115.
[4] space passes China, Xu Yongyong. and non parametric method is estimated ROC area under curve [J]. Chinese health statistics, 1999,16 (4): 241-244.
[5]Richard?E.Kerber，Chair?MD，Lance?B.Becker，et?al.AutomaticExternal?Defibrillators?for?Public?Access?Defibrillation：Recommendationsfor?Specifying?and?Reporting?Arrhythmia?Analysis?Algorithm?Performance，Incorporating?New?Waveforms，and?Enhancing?Safety[J].Circulation，1997，95(6)：1677-1682.
Claims (3)
1. but improved Electrical Cardioversion rhythm of the heart identification instrument based on grid projection distribution dispersion is characterized in that comprising:
S1. pretreatment module: the electrocardiosignal that collects is carried out pretreatment;
S2. whether the identification module of the asystole rhythm of the heart: being used to differentiate electrocardiosignal is the asystole rhythm of the heart;
S3. calculate the module of grid projection distribution dispersion;
But Electrical Cardioversion rhythm of the heart discrimination module S4.: whether the grid projection distribution dispersion that calculates according to the module of calculating grid projection distribution dispersion is differentiated is can not the Electrical Cardioversion rhythm of the heart, if grid projection distribution dispersion＞=threshold value, then being judged to can not the Electrical Cardioversion rhythm of the heart; If grid projection distribution dispersion＜threshold value, but then be judged to the Electrical Cardioversion rhythm of the heart.
2. instrument according to claim 1 is characterized in that, described pretreatment module comprises:
S11. one 5 rank moving average filter is used for filter away high frequency noise;
S12. the high pass filter that cut-off frequency is 1Hz is used to suppress baseline drift;
S13. the Butterworth low pass filter that cut-off frequency is 30Hz is used for the irrelevant radio-frequency component of further filtering.
3. instrument according to claim 1 and 2 is characterized in that, the module of described calculating grid projection distribution dispersion comprises:
S31. grizzly bar is cut apart submodule, and this submodule is used for one section electrocardiogram (ECG) data is divided into segment by identical interval, and each segment is called a grizzly bar, and each interval is called grill width;
S32. grid projection calculating sub module, this submodule is used for the amplitude range that calculating ECG covers in each grizzly bar, be exactly the projection on grizzly bar y axle;
S33. grid projection value distribution statistics submodule, this submodule is used for being divided into several zones between the maximum and minima of all grid projections, and statistics drops on the quantity of the grid projection in each zone;
S34. grid projection distribution dispersion calculating sub module, this submodule is used to calculate grid projection distribution dispersion.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2009100451537A CN101461710B (en) | 2009-01-12 | 2009-01-12 | Shockable rhythm recognition algorithm based on grid projection distribution dispersion |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2009100451537A CN101461710B (en) | 2009-01-12 | 2009-01-12 | Shockable rhythm recognition algorithm based on grid projection distribution dispersion |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101461710A CN101461710A (en) | 2009-06-24 |
| CN101461710B true CN101461710B (en) | 2010-12-29 |
Family
ID=40802653
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2009100451537A Expired - Fee Related CN101461710B (en) | 2009-01-12 | 2009-01-12 | Shockable rhythm recognition algorithm based on grid projection distribution dispersion |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101461710B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107376122A (en) * | 2017-07-14 | 2017-11-24 | 上海救要救信息科技有限公司 | A kind of automated external defibrillator system |
| CN110384482A (en) * | 2019-06-26 | 2019-10-29 | 广州视源电子科技股份有限公司 | Electrocardiosignal classification method and device, computer equipment and storage medium |
| CN110226917A (en) * | 2019-06-26 | 2019-09-13 | 广州视源电子科技股份有限公司 | Electrocardiosignal type detection method and device, computer equipment and storage medium |
| CN110226919B (en) * | 2019-06-26 | 2022-05-03 | 广州视源电子科技股份有限公司 | Electrocardiosignal type detection method and device, computer equipment and storage medium |
| CN110226918B (en) * | 2019-06-26 | 2022-06-28 | 广州视源电子科技股份有限公司 | Electrocardiosignal type detection method and device, computer equipment and storage medium |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1458852A (en) * | 2001-03-13 | 2003-11-26 | 菲利浦电子北美公司 | Interactive method of performing cardipulmonary resuscitation with minimal delay to defibrillation shocks |
| CN101203267A (en) * | 2005-06-23 | 2008-06-18 | 皇家飞利浦电子股份有限公司 | Defibrillator with automatic shock first/CPR first algorithm |
-
2009
- 2009-01-12 CN CN2009100451537A patent/CN101461710B/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1458852A (en) * | 2001-03-13 | 2003-11-26 | 菲利浦电子北美公司 | Interactive method of performing cardipulmonary resuscitation with minimal delay to defibrillation shocks |
| CN101203267A (en) * | 2005-06-23 | 2008-06-18 | 皇家飞利浦电子股份有限公司 | Defibrillator with automatic shock first/CPR first algorithm |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101461710A (en) | 2009-06-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101461711A (en) | Shockable rhythm recognition algorithm based on standard deviation of standard slope absolute value | |
| Hammad et al. | Detection of abnormal heart conditions based on characteristics of ECG signals | |
| CN104382590B (en) | What a kind of combination electrocardio time and frequency domain characteristics were analyzed can Electrical Cardioversion rhythm of the heart automatic identification and classifying method | |
| Tripathy et al. | Detection of shockable ventricular arrhythmia using variational mode decomposition | |
| CN101461709B (en) | Shockable rhythm recognition instrument | |
| Addison et al. | Evaluating arrhythmias in ECG signals using wavelet transforms | |
| Murukesan et al. | Machine learning approach for sudden cardiac arrest prediction based on optimal heart rate variability features | |
| US20180325403A1 (en) | Non-invasive method and system for characterizing cardiovascular systems for all-cause mortality and sudden cardiac death risk | |
| CN101461710B (en) | Shockable rhythm recognition algorithm based on grid projection distribution dispersion | |
| Romero et al. | ECG frequency domain features extraction: A new characteristic for arrhythmias classification | |
| Malek et al. | Automated detection of premature ventricular contraction in ECG signals using enhanced template matching algorithm | |
| Sadr et al. | A low-complexity algorithm for detection of atrial fibrillation using an ECG | |
| Murugappan et al. | Time domain features based sudden cardiac arrest prediction using machine learning algorithms | |
| Parsi et al. | Prediction of sudden cardiac death in implantable cardioverter defibrillators: a review and comparative study of heart rate variability features | |
| Hotradat et al. | Empirical mode decomposition based ECG features in classifying and tracking ventricular arrhythmias | |
| Balasundaram et al. | Wavelet-based features for characterizing ventricular arrhythmias in optimizing treatment options | |
| Dliou et al. | Abnormal ECG signals analysis using non-parametric time–frequency techniques | |
| Riasi et al. | Prediction of ventricular tachycardia using morphological features of ECG signal | |
| CN101474069B (en) | Improved shockable cardioversion identifying instrument | |
| Othman et al. | A new semantic mining approach for detecting ventricular tachycardia and ventricular fibrillation | |
| Murthy et al. | ECG signal denoising and ischemic event feature extraction using Daubechies wavelets | |
| Khadra et al. | A new quantitative analysis technique for cardiac arrhythmia using bispectrum and bicoherency | |
| Nemirko et al. | The Comparison of Algorithms for Life-threatening Cardiac Arrhythmias Recognition. | |
| Afatmirni et al. | Predicting refibrillation from pre-shock waveforms in optimizing cardiac resuscitation | |
| Zhang et al. | A shockable rhythm detection algorithm for automatic external defibrillators by combining a slope variability analyzer with a band-pass digital filter |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20101229 Termination date: 20150112 |
|
| EXPY | Termination of patent right or utility model |