CN103584852A - Personalized electrocardiogram intelligent auxiliary diagnosis device and method - Google Patents
Personalized electrocardiogram intelligent auxiliary diagnosis device and method Download PDFInfo
- Publication number
- CN103584852A CN103584852A CN201210290454.8A CN201210290454A CN103584852A CN 103584852 A CN103584852 A CN 103584852A CN 201210290454 A CN201210290454 A CN 201210290454A CN 103584852 A CN103584852 A CN 103584852A
- Authority
- CN
- China
- Prior art keywords
- electrocardiogram
- heart state
- vector
- grader
- personalized
- 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.)
- Pending
Links
Images
Landscapes
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention discloses a personalized electrocardiogram intelligent auxiliary diagnosis device which comprises a feature vector extraction module used for extracting a to-be-trained electrocardiogram feature vector, a heart state classifier generation module used for training the to-be-trained feature vector to generate a heart state classifier, a user identification module used for obtaining user identity label, a heart state classifier storage module used for storing a heart state classifier of a corresponding user based on the user identity label, and a diagnosis module used for obtaining the heart state classifier of the corresponding user based on the user identity label and diagnosing a to-be-diagnosed electrocardiogram feature vector through the heart state classifier to obtain heart state classification, wherein the feature vector extraction module is also used for extracting the to-be-diagnosed electrocardiogram feature vector and a feedback electrocardiogram feature vector, and the heart state classifier generation module is also used for training the feedback electrocardiogram feature vector and updating the heart state classifier. According to the personalized electrocardiogram intelligent auxiliary diagnosis device and method, individual diagnosis can be performed on different individuals, and reference is provided for diagnosis of doctors.
Description
Technical field
The present invention relates to electrocardiographic diagnosis field, relate in particular to a kind of personalized electrocardiogram intelligent auxiliary diagnosis apparatus and method.
Background technology
Electrocardiogram has important reference value for research and the cardiopathic diagnosis of heart basic function.Yet due to the congenital difference of individuality, even the in the situation that of health of heart, the parameters index of the single heartbeat of each normal person's electrocardiogram is different.Electrocardiographic diagnosis systems more of the prior art directly utilize previously electrocardiogram to carry out learning classification, but ignored individual variation for Electrocardiographic impact, make to have adopted identical index to diagnose for every patient, such electrocardiographic diagnosis method is science not.
Therefore, design and a kind ofly can to Different Individual, implement the electrocardiogram diagnosis device of personalized diagnosis, will effectively overcome the diagnosis index that exists in prior art not because of the different defect of individual difference, contribute to improve the accuracy rate of electrocardiographic diagnosis.
Summary of the invention
The present invention is intended to solve above-mentioned problems of the prior art, proposes a kind of personalized electrocardiogram intelligent auxiliary diagnosis apparatus and method.On the one hand, the personalized electrocardiogram intelligent auxiliary diagnosis device that the present invention proposes comprises: characteristic vector extraction module, heart state grader generation module, user identification module, heart state grader memory module and diagnostic module.Wherein,
The characteristic vector of some single heartbeats in electrocardiogram to be instructed described in characteristic vector extraction module reception user electrocardiogram to be instructed extraction, heart state grader generation module is trained Characteristics of electrocardiogram vector described to be instructed, generate heart state grader, wherein, described characteristic vector extraction module also receives user's follow-up electrocardiogram and extracts the characteristic vector of some single heartbeats in described follow-up electrocardiogram, described heart state grader, according to the described follow-up Characteristics of electrocardiogram vector of input, is exported heart state classification corresponding to described follow-up Characteristics of electrocardiogram vector; User identification module obtains User Identity; Heart state grader memory module is according to the described heart state grader of described User Identity storage relative users; Diagnostic module obtains the described heart state grader of relative users according to described User Identity, and by described heart state grader, described follow-up Characteristics of electrocardiogram vector is diagnosed, draw heart state classification corresponding to described follow-up Characteristics of electrocardiogram vector; Wherein, described characteristic vector extraction module also receives feedback electrocardiogram and extracts the characteristic vector of some single heartbeats in described feedback electrocardiogram, described heart state grader generation module is also trained described feedback Characteristics of electrocardiogram vector, and upgrades described heart state grader.
On the other hand, the personalized electrocardiogram intelligent auxiliary diagnosis device that the present invention proposes comprises: the characteristic vector of some single heartbeats in electrocardiogram to be instructed described in reception user electrocardiogram to be instructed extraction; Characteristics of electrocardiogram vector described to be instructed is trained, generate heart state grader, described heart state grader, for according to the described follow-up Characteristics of electrocardiogram vector of input, is exported heart state classification corresponding to described follow-up Characteristics of electrocardiogram vector; Obtain User Identity; According to the described heart state grader of described User Identity storage relative users; According to described User Identity, obtain the described heart state grader of relative users, and by described heart state grader, described follow-up Characteristics of electrocardiogram vector is diagnosed, draw heart state classification corresponding to described follow-up Characteristics of electrocardiogram vector; Receive the characteristic vector of feeding back electrocardiogram and extracting some single heartbeats in described feedback electrocardiogram; Described feedback Characteristics of electrocardiogram vector is trained, and upgrade described heart state grader.
The personalized electrocardiogram intelligent auxiliary diagnosis apparatus and method of the embodiment of the present invention are by resident's the past electrocardiogram, training generates the heart state grader that belongs to this resident, by described heart state grader, this resident's follow-up electrocardiogram is diagnosed, realized the personalization diagnosis to different diseased individuals, contribute to improve diagnosis, thereby for diagnosis provides reference, contribute to alleviate doctor's medical burden.
Accompanying drawing explanation
Below in conjunction with accompanying drawing, the present invention is described in detail, wherein:
Fig. 1 is the personalized electrocardiogram intelligent auxiliary diagnosis structure drawing of device of one embodiment of the invention;
Fig. 2 is the personalized electrocardiogram intelligent auxiliary diagnosis structure drawing of device of another embodiment of the present invention.
Fig. 3 is the personalized electrocardiogram intelligent auxiliary diagnosis method flow diagram of one embodiment of the invention;
Fig. 4 is the personalized electrocardiogram intelligent auxiliary diagnosis method flow diagram of another embodiment of the present invention;
The specific embodiment
Below by drawings and Examples, technical solution of the present invention is described in further detail.
One aspect of the present invention proposes a kind of personalized electrocardiogram intelligent auxiliary diagnosis device, the personalized electrocardiogram intelligent auxiliary diagnosis structure drawing of device that Fig. 1 is one embodiment of the invention.
In one embodiment, first to the some residents in a certain area, collect respectively the electrocardiogram of its 5 years the pasts, often occupy the 10 width electrocardiograms of getting elected, described electrocardiogram is made a definite diagnosis by doctor, and the result of making a definite diagnosis is a kind of in following 10 kinds of heart state classifications: normal sinus heart rate, room pass to retardance, supraventricular tachycardia, ventricular tachycardia, nodal rhythm, ventricular fibrillation, left focus ventricular premature contraction, right focus ventricular premature contraction, left bundle branch block and right bundle branch block.Wherein, normal sinus heart rate is heart normal condition, and all the other 9 kinds is heart abnormality state.The described 10 width electrocardiograms that often occupy the people are electrocardiogram to be instructed.In the present embodiment, by the category label that the category label of normal sinus heart rate is designated as 1, room passes to retardance be designated as 2, the category label of supraventricular tachycardia is designated as 3, the category label of ventricular tachycardia is designated as 4, the category label of nodal rhythm is designated as 5, the category label of ventricular fibrillation is designated as 6, the category label of left focus ventricular premature contraction is designated as 7, the category label of right focus ventricular premature contraction is designated as 8, the category label of left bundle branch block is designated as 9, the category label of right bundle branch block is designated as 10, trains waiting.
Characteristic vector extraction module 100 is for electrocardiogram to be instructed described in receiving and extract the characteristic vector of described a certain resident's the some single heartbeats of electrocardiogram to be instructed.In the present embodiment, described characteristic vector is ten dimensional feature vectors, comprising: the ratio of single heartbeat effective time, QRS interval, QT interval, PR interval, P peak point, T peak point, P wave duration, T wave duration, p wave duration and PR interval.Every width electrocardiogram consists of a plurality of single heartbeats, and each single heartbeat accounts for certain cycle, and described characteristic vector is the build-in attribute of single heartbeat, is also the important indicator of the described 10 kinds of heart state classifications of reflection.In described characteristic vector, some dimension component values is abnormal, has reflected that heart is in corresponding abnormality.
In the present embodiment, the 10 width electrocardiograms for a certain resident, it is example that the described 10 width electrocardiograms of take have comprised all 10 kinds of heart state classifications, and described characteristic vector extraction module 100 extracts ten dimensional feature vectors to be instructed under 100 normal sinus heart rate states, is designated as X
1, X
2..., X
100, wherein, X
ibe ten dimensional feature vectors, category label y
1=1; Extract 100 rooms and pass to ten dimensional feature vectors to be instructed under retardance state, be designated as X
101, X
102..., X
200, wherein, category label y2=2; The like, extract ten dimensional feature vectors to be instructed under 100 right bundle branch block states, be designated as X
901, X
902..., X
1000, wherein, category label y
10=10.
In the present embodiment, described characteristic vector extraction module 100 is also for receiving resident's follow-up electrocardiogram and extracting the characteristic vector of the some single heartbeats of described follow-up electrocardiogram.
Heart state grader generation module 200, for described a certain resident Characteristics of electrocardiogram vector to be instructed is trained, generates heart state grader.In the present embodiment, adopt support vector machine (SVM) to characteristic vector { (X described to be instructed
1, y
1), (X
2, y
1) ..., (X
100, y
1), (X
101, y
2), (X
102, y
2) ..., (X
200, y
2) ..., (X
601, y
10), (X
602, y
10) ..., (X
700, y
10) train, generating heart state grader, described heart state grader can be used for input sample X
tclassify, to determine X
tbelong to which classification in above-mentioned 10 kinds of heart state classifications, wherein, X
tfor Electrocardiographic ten dimensional feature vectors of described follow-up.
Wherein, characteristic vector { (X to be instructed described in 200 pairs of heart state grader generation modules
1, y
1), (X
2, y
1) ..., (X
100, y
1), (X
10i, y
2), (X
102, y
2) ..., (X
200, y
2) ..., (X
601, y
10), (X
602, y
10) ..., (X
700, y
10) to train the algorithm adopting can be OAO-SVMs algorithm, described OAO-SVMs algorithm is by several two classification graders of structure, again by the synthetic very class grader of described several two classification set of classifiers, described very class grader is for classifying to described input sample XT, to determine XT belongs to which classification in described 10 kinds of heart state classifications.
In the present embodiment, heart state grader corresponding to different residents.
Described above is personalized according to an embodiment of the invention electrocardiogram intelligent auxiliary diagnosis device, resident's electrocardiogram to be instructed is trained to the operation principle of the heart state grader that obtains described resident, will be described further the Electrocardiographic diagnosis of resident's follow-up in above-described embodiment below.
Characteristic vector extraction module 100 is for receiving described follow-up electrocardiogram and extracting the characteristic vector of the some single heartbeats of described follow-up electrocardiogram.
Described user identification module 400 obtains described follow-up resident's identify label.If this follow-up resident once used described personalized electrocardiogram intelligent auxiliary diagnosis device to train its past electrocardiogram, described diagnostic module 500 obtains described follow-up resident's heart state grader in described heart state grader memory module 300 according to described follow-up resident's identify label, and by described heart state grader, described follow-up Characteristics of electrocardiogram vector is diagnosed, show described follow-up Characteristics of electrocardiogram vector is corresponding to which kind of classification in described 10 kinds of heart state classifications.
Described above is personalized according to an embodiment of the invention electrocardiogram intelligent auxiliary diagnosis device, the operation principle that resident's follow-up electrocardiogram is diagnosed, below by above-described embodiment according to feedback electrocardiogram the renewal of described heart state grader is described further.
In support vector machine theory, its classification accuracy of the grader obtaining after training is directly proportional to the quantity of sample to be instructed, yet the quantity of sample to be instructed is huger, and training cost is higher.In the present embodiment, the quantity of each heart state classification correspondence characteristic vector to be instructed is 100, due to the limited amount of characteristic vector described to be instructed, described personalized electrocardiogram intelligent auxiliary diagnosis device can not guarantee the entirely accurate to described follow-up electrocardiographic diagnosis.
Therefore, at the initial stage of using described personalized electrocardiogram intelligent auxiliary diagnosis device to diagnose described follow-up electrocardiogram, doctor should diagnose described follow-up electrocardiogram simultaneously, when the diagnostic result of described personalized electrocardiogram intelligent auxiliary diagnosis device and diagnosis result are when inconsistent, should get diagnosis result is last diagnostic result, and give described personalized electrocardiogram intelligent auxiliary diagnosis device by described diagnosis result feedback, described heart state grader is upgraded, to improve the diagnosis of described personalized electrocardiogram intelligent auxiliary diagnosis device.After the accuracy rate of diagnosis of the described personalized electrocardiogram intelligent auxiliary diagnosis device after constantly updating reaches certain threshold value, can not need diagnosis result to upgrade the feedback of described heart state grader, thereby provide reference for diagnosis.
Particularly, the some follow-up Characteristics of electrocardiogram vector Xs of described personalized electrocardiogram intelligent auxiliary diagnosis device to a certain resident
t1, X
t2... the result of diagnosis is normal sinus heart rate, and the follow-up Characteristics of electrocardiogram vector X of doctor to this resident
t1, X
t2... diagnostic result be nodal rhythm, so now should get nodal rhythm is last diagnostic result, the category label of nodal rhythm is 5, then by { (X
t1, 5), (X
t2, 5) ... } as feeding back signal to be instructed, input 200 pairs of described heart state graders renewal training of described heart grader generation module, the heart state grader after being upgraded.
Heart state grader after heart state grader memory module 300 is upgraded according to this resident of above-mentioned residential identity sign storage, for the electrocardiographic diagnosis to this resident from now on.
Preferably, as described in Figure 2, described personalized electrocardiogram intelligent auxiliary diagnosis device also comprises ecg measurement module 600 and electrocardiogram memory module 700.
Described ecg measurement module 600 is for electrocardiogram to be instructed or described follow-up electrocardiogram described in measuring.Described electrocardiogram memory module 700, for having the processor of storage medium, includes but not limited to: personal computer, server.The residential identity sign of described ecg measurement module 600 for obtaining according to described user identification module 400, described in inciting somebody to action, electrocardiogram to be instructed or described follow-up electrocardiogram are stored in the corresponding memory space of described resident.
The operation principle of all modules in Fig. 2 except ecg measurement module 600 and electrocardiogram memory module 700 is described in detail in to the introduction of Fig. 1, repeats no more herein.
The present invention proposes a kind of personalized electrocardiogram intelligent auxiliary diagnosis method on the other hand, the personalized electrocardiogram intelligent auxiliary diagnosis method flow diagram that Fig. 3 is one embodiment of the invention.
In one embodiment, execution step S100 is before first to the some residents in a certain area, collect respectively the electrocardiogram of its 5 years the pasts, often occupy the 10 width electrocardiograms of getting elected, described electrocardiogram is made a definite diagnosis by doctor, and the result of making a definite diagnosis is a kind of in following 10 kinds of heart state classifications: normal sinus heart rate, room pass to retardance, supraventricular tachycardia, ventricular tachycardia, nodal rhythm, ventricular fibrillation, left focus ventricular premature contraction, right focus ventricular premature contraction, left bundle branch block and right bundle branch block.Wherein, normal sinus heart rate is heart normal condition, and all the other 9 kinds is heart abnormality state.The described 10 width electrocardiograms that often occupy the people are electrocardiogram to be instructed.In the present embodiment, by the category label that the category label of normal sinus heart rate is designated as 1, room passes to retardance be designated as 2, the category label of supraventricular tachycardia is designated as 3, the category label of ventricular tachycardia is designated as 4, the category label of nodal rhythm is designated as 5, the category label of ventricular fibrillation is designated as 6, the category label of left focus ventricular premature contraction is designated as 7, the category label of right focus ventricular premature contraction is designated as 8, the category label of left bundle branch block is designated as 9, the category label of right bundle branch block is designated as 10, trains waiting.
In step S100, electrocardiogram to be instructed extract the characteristic vector of some single heartbeats in described a certain resident's electrocardiogram to be instructed described in reception.In the present embodiment, described characteristic vector is ten dimensional feature vectors, comprising: the ratio of single heartbeat effective time, QRS interval, QT interval, PR interval, P peak point, T peak point, P wave duration, T wave duration, p wave duration and PR interval.Every width electrocardiogram consists of a plurality of single heartbeats, and each single heartbeat accounts for certain cycle, and described characteristic vector is the build-in attribute of single heartbeat, is also the important indicator of the described 10 kinds of heart state classifications of reflection.In described characteristic vector, some dimension component values is abnormal, has reflected that heart is in corresponding abnormality.
In the present embodiment, for a certain resident's 10 width electrocardiograms, it is example that the described 10 width electrocardiograms of take have comprised all 10 kinds of heart state classifications, in step S100, extracts ten dimensional feature vectors to be instructed under 100 normal sinus heart rate states, is designated as X
1, X
2..., X
100, wherein, X
ibe ten dimensional feature vectors, category label y
1=1; Extract 100 rooms and pass to ten dimensional feature vectors to be instructed under retardance state, be designated as X
101, X
102..., X
200, wherein, category label y
2=2; The like, extract ten dimensional feature vectors to be instructed under 100 right bundle branch block states, be designated as X
901, X
902..., X
1000, wherein, category label y
10=10.
In step S200, described a certain resident Characteristics of electrocardiogram vector to be instructed is trained, generate heart state grader.In the present embodiment, adopt support vector machine (SVM) to characteristic vector { (X described to be instructed
1, y
1), (X
2, y
1) ..., (X
100, y
1), (X
101, y
2), (X
102, y
2) ..., (X
200, y
2) ..., (X
601, y
10), (X
602, y
10) ..., (X
700, y
10) train, generating heart state grader, described heart state grader can be used for input sample X
tclassify, to determine X
tbelong to which classification in above-mentioned 10 kinds of heart state classifications, wherein, X
tfor Electrocardiographic ten dimensional feature vectors of described follow-up.
Wherein, to characteristic vector { (X described to be instructed
1, y
1), (X
2, y
1) ..., (X
100, y
1), (X
101, y
2), (X
102, y
2) ..., (X
200, y
2) ..., (X
601, y
10), (X
602, y
10) ..., (X
700, y
10) to train the algorithm adopting can be OAO-SVMs algorithm, described OAO-SVMs algorithm is by several two classification graders of structure, by the synthetic very class grader of described several two classification set of classifiers, described very class grader is used for described input sample X again
tclassify, to determine X
tbelong to which classification in described 10 kinds of heart state classifications.
In step S300, obtain User Identity, the affiliated resident of characteristic vector to be instructed described in described User Identity is used for identifying, includes but not limited to: resident's name, resident identification card number.
In step S400, according to the described heart state grader of described User Identity storage relative users.Storage tool used is the processor with storage medium, includes but not limited to: personal computer, server.Storage tool used opens up according to the User Identity of input the memory space that belongs to this user in its storage medium, and will after training, obtain being stored in described memory space corresponding to this user's heart state grader, for to the Electrocardiographic diagnosis of resident's follow-up.
In the present embodiment, heart state grader corresponding to different residents.
In step S500, the follow-up resident's who has stored described in obtaining according to described follow-up resident's identify label heart state grader, and by described heart state grader, described follow-up Characteristics of electrocardiogram vector is diagnosed, show described follow-up Characteristics of electrocardiogram vector is corresponding to which kind of classification in described 10 kinds of heart state classifications.
In support vector machine theory, its classification accuracy of the grader obtaining after training is directly proportional to the quantity of sample to be instructed, yet the quantity of sample to be instructed is huger, and training cost is higher.In the present embodiment, the quantity of each heart state classification correspondence characteristic vector to be instructed is 100, due to the limited amount of characteristic vector described to be instructed, described personalized electrocardiogram intelligent auxiliary diagnosis method can not guarantee the entirely accurate to described follow-up electrocardiographic diagnosis.
Therefore, at the initial stage of using described personalized electrocardiogram intelligent auxiliary diagnosis method to diagnose described follow-up electrocardiogram, doctor should diagnose described follow-up electrocardiogram simultaneously, when the diagnostic result of described personalized electrocardiogram intelligent auxiliary diagnosis method and diagnosis result are when inconsistent, should get diagnosis result is last diagnostic result, and described diagnosis result is fed back, described heart state grader is upgraded, to improve the diagnosis of described personalized electrocardiogram intelligent auxiliary diagnosis method.After the accuracy rate of diagnosis of the described personalized electrocardiogram intelligent auxiliary diagnosis method after constantly updating reaches certain threshold value, can not need diagnosis result to upgrade the feedback of described heart state grader, thereby provide reference for diagnosis.
In step S600, receive the characteristic vector of feeding back electrocardiogram and extracting some single heartbeats in described feedback electrocardiogram.Particularly, use the some follow-up Characteristics of electrocardiogram vector Xs of described personalized electrocardiogram intelligent auxiliary diagnosis method to a certain resident
t1, X
t2... the result of diagnosis is normal sinus heart rate, and the follow-up Characteristics of electrocardiogram vector X of doctor to this resident
t1, X
t2... diagnostic result be nodal rhythm, so now should get nodal rhythm is last diagnostic result, the category label of nodal rhythm is 5, then by { (X
t1, 5), (X
t2, 5) ... } as feeding back signal to be instructed, described heart state grader is upgraded to training.In step S700, described feedback characteristic vector to be trained, the heart state grader after being upgraded, for the electrocardiographic diagnosis to this resident from now on.
Preferably, as described in Figure 4, described personalized electrocardiogram intelligent auxiliary diagnosis method also comprises step S800 and step S900.Particularly, in step S800, measure the electrocardiogram that described user's electrocardio obtains described user.In step S900, according to described User Identity, store described user's electrocardiogram.Described electrocardiogram comprises electrocardiogram to be instructed and follow-up electrocardiogram, and the training after step S900 and diagnostic procedure are described in detail in to the introduction of Fig. 3, repeat no more herein.
The personalized electrocardiogram intelligent auxiliary diagnosis apparatus and method of the embodiment of the present invention are by resident's the past electrocardiogram, training generates the heart state grader that belongs to this resident, by described heart state grader, this resident's follow-up electrocardiogram is diagnosed, realized the personalization diagnosis to different diseased individuals, contribute to improve diagnosis, thereby for diagnosis provides reference, contribute to alleviate doctor's medical burden.
Although the present invention is described with reference to current preferred embodiments; but those skilled in the art will be understood that; above-mentioned preferred embodiments is only used for illustrating the present invention; not be used for limiting protection scope of the present invention; any within the spirit and principles in the present invention scope; any modification of doing, equivalence replacement, improvement etc., within all should being included in the scope of the present invention.
Claims (10)
1. a personalized electrocardiogram intelligent auxiliary diagnosis device, for user's electrocardiogram is diagnosed, is characterized in that, comprising:
Characteristic vector extraction module, for receive user's electrocardiogram to be instructed and extract described in the characteristic vector of the some single heartbeats of electrocardiogram to be instructed;
Heart state grader generation module, for Characteristics of electrocardiogram vector described to be instructed is trained, generate heart state grader, wherein, described characteristic vector extraction module is also for receiving user's follow-up electrocardiogram and extracting the characteristic vector of the some single heartbeats of described follow-up electrocardiogram, described heart state grader, for according to the described follow-up Characteristics of electrocardiogram vector of input, is exported heart state classification corresponding to described follow-up Characteristics of electrocardiogram vector;
User identification module, for obtaining User Identity;
Heart state grader memory module, for storing the described heart state grader of relative users according to described User Identity;
Diagnostic module, for obtain the described heart state grader of relative users according to described User Identity, and by described heart state grader, described follow-up Characteristics of electrocardiogram vector is diagnosed, draw heart state classification corresponding to described follow-up Characteristics of electrocardiogram vector;
Wherein, described characteristic vector extraction module is also for receiving the characteristic vector of feeding back electrocardiogram and extracting the some single heartbeats of described feedback electrocardiogram, described heart state grader generation module is also for described feedback Characteristics of electrocardiogram vector is trained, and upgrades described heart state grader.
2. personalized electrocardiogram intelligent auxiliary diagnosis device according to claim 1, it is characterized in that, described heart state classification comprises: normal sinus heart rate, room pass to retardance, supraventricular tachycardia, ventricular tachycardia, nodal rhythm, ventricular fibrillation, left focus ventricular premature contraction, right focus ventricular premature contraction, left bundle branch block and right bundle branch block.
3. personalized electrocardiogram intelligent auxiliary diagnosis device according to claim 1, it is characterized in that, described characteristic vector is ten dimensional feature vectors, comprising: the ratio of single heartbeat effective time, QRS interval, QT interval, PR interval, P peak point, T peak point, P wave duration, T wave duration, p wave duration and PR interval.
4. personalized electrocardiogram intelligent auxiliary diagnosis device according to claim 1, is characterized in that, described heart state grader is support vector machine (SVM) grader.
5. personalized electrocardiogram intelligent auxiliary diagnosis device according to claim 1, is characterized in that, also comprises:
Ecg measurement module, obtains described user's electrocardiogram for measuring described user's electrocardio;
Electrocardiogram memory module, for storing described user's electrocardiogram according to described User Identity.
6. a personalized electrocardiogram intelligent auxiliary diagnosis method, for user's electrocardiogram is diagnosed, is characterized in that, comprising:
The characteristic vector of some single heartbeats in electrocardiogram to be instructed described in reception user electrocardiogram to be instructed extraction;
Characteristics of electrocardiogram vector described to be instructed is trained, generate heart state grader, described heart state grader, for according to the described follow-up Characteristics of electrocardiogram vector of input, is exported heart state classification corresponding to described follow-up Characteristics of electrocardiogram vector;
Obtain User Identity;
According to the described heart state grader of described User Identity storage relative users;
According to described User Identity, obtain the described heart state grader of relative users, and by described heart state grader, described follow-up Characteristics of electrocardiogram vector is diagnosed, draw heart state classification corresponding to described follow-up Characteristics of electrocardiogram vector;
Receive the characteristic vector of feeding back electrocardiogram and extracting some single heartbeats in described feedback electrocardiogram;
Described feedback Characteristics of electrocardiogram vector is trained, and upgrade described heart state grader.
7. personalized electrocardiogram intelligent auxiliary diagnosis method according to claim 6, it is characterized in that, described heart state classification comprises: normal sinus heart rate, room pass to retardance, supraventricular tachycardia, ventricular tachycardia, nodal rhythm, ventricular fibrillation, left focus ventricular premature contraction, right focus ventricular premature contraction, left bundle branch block and right bundle branch block.
8. personalized electrocardiogram intelligent auxiliary diagnosis method according to claim 6, it is characterized in that, described characteristic vector is ten dimensional feature vectors, comprising: the ratio of single heartbeat effective time, QRS interval, QT interval, PR interval, P peak point, T peak point, P wave duration, T wave duration, p wave duration and PR interval.
9. personalized electrocardiogram intelligent auxiliary diagnosis method according to claim 6, is characterized in that, described heart state grader is support vector machine (SVM) grader.
10. personalized electrocardiogram intelligent auxiliary diagnosis method according to claim 6, is characterized in that, also comprises:
Measure the electrocardiogram that described user's electrocardio obtains described user; According to described User Identity, store described user's electrocardiogram.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210290454.8A CN103584852A (en) | 2012-08-15 | 2012-08-15 | Personalized electrocardiogram intelligent auxiliary diagnosis device and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210290454.8A CN103584852A (en) | 2012-08-15 | 2012-08-15 | Personalized electrocardiogram intelligent auxiliary diagnosis device and method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103584852A true CN103584852A (en) | 2014-02-19 |
Family
ID=50075330
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210290454.8A Pending CN103584852A (en) | 2012-08-15 | 2012-08-15 | Personalized electrocardiogram intelligent auxiliary diagnosis device and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103584852A (en) |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104382582A (en) * | 2014-11-10 | 2015-03-04 | 哈尔滨医科大学 | Device for classifying dynamic electrocardio data |
| CN104398254A (en) * | 2014-11-14 | 2015-03-11 | 中国科学院深圳先进技术研究院 | Electrocardiogram analyzing system, electrocardiogram analyzing equipment and electrocardiogram predication model acquisition equipment |
| CN104523266A (en) * | 2015-01-07 | 2015-04-22 | 河北大学 | Automatic classification method for electrocardiogram signals |
| CN105468951A (en) * | 2015-11-17 | 2016-04-06 | 安徽华米信息科技有限公司 | Method and device for identity recognition through electrocardiographic feature and wearable device |
| CN105769171A (en) * | 2014-12-26 | 2016-07-20 | 中国移动通信集团公司 | Arrhythmia detection method and device |
| WO2017084546A1 (en) * | 2015-11-17 | 2017-05-26 | 安徽华米信息科技有限公司 | Wearable device-based user interest information determination method, device and wearable device |
| CN106901724A (en) * | 2017-04-26 | 2017-06-30 | 河南省中医院（河南中医药大学第二附属医院） | Electrocardiogram testing device and testing method |
| CN106963481A (en) * | 2016-01-14 | 2017-07-21 | 韦伯斯特生物官能(以色列)有限公司 | Focal source is recognized during atrial fibrillation |
| CN107451417A (en) * | 2017-09-08 | 2017-12-08 | 北京蓬阳丰业医疗设备有限公司 | Dynamic ECG analysis intelligent diagnosis system and method |
| WO2018184202A1 (en) * | 2017-04-07 | 2018-10-11 | 北京悦琦创通科技有限公司 | Inference engine training method and device, upgrade method, device and storage medium |
| CN109222957A (en) * | 2018-08-29 | 2019-01-18 | 西安电子科技大学 | A kind of automation ECG measuring system |
| CN111356401A (en) * | 2017-11-29 | 2020-06-30 | 威里利生命科学有限责任公司 | Continuous detection and monitoring of cardiac arrhythmias using wearable sensors and cloud residency analysis |
| CN113679347A (en) * | 2021-08-24 | 2021-11-23 | 安徽相王医疗健康股份有限公司 | Diagnosis and treatment effect evaluation method based on data statistics |
| US11806152B2 (en) | 2016-01-14 | 2023-11-07 | Biosense Webster (Israel), Ltd. | Non-overlapping loop-type or spline-type catheter to determine activation source direction and activation source type |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1605974A (en) * | 2003-10-06 | 2005-04-13 | 索尼株式会社 | Information recording device and information recording method |
| CN101908152A (en) * | 2010-06-11 | 2010-12-08 | 电子科技大学 | Customization classifier-based eye state identification method |
| WO2011043978A2 (en) * | 2009-10-05 | 2011-04-14 | Elc Management Llc | Computer-aided diagnostic systems and methods for determining skin compositions based on traditional chinese medicinal (tcm) principles |
| CN102165453A (en) * | 2008-09-26 | 2011-08-24 | 皇家飞利浦电子股份有限公司 | System and method for fusing clinical and image features for computer-aided diagnosis |
-
2012
- 2012-08-15 CN CN201210290454.8A patent/CN103584852A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1605974A (en) * | 2003-10-06 | 2005-04-13 | 索尼株式会社 | Information recording device and information recording method |
| CN102165453A (en) * | 2008-09-26 | 2011-08-24 | 皇家飞利浦电子股份有限公司 | System and method for fusing clinical and image features for computer-aided diagnosis |
| WO2011043978A2 (en) * | 2009-10-05 | 2011-04-14 | Elc Management Llc | Computer-aided diagnostic systems and methods for determining skin compositions based on traditional chinese medicinal (tcm) principles |
| CN102421357A (en) * | 2009-10-05 | 2012-04-18 | Elc管理有限责任公司 | Computer-aided diagnostic systems and methods for determining skin compositions based on traditional chinese medicinal (tcm) principles |
| CN101908152A (en) * | 2010-06-11 | 2010-12-08 | 电子科技大学 | Customization classifier-based eye state identification method |
Non-Patent Citations (3)
| Title |
|---|
| 周珂,彭宏: "支持向量机在心电图分类诊断中的应用", 《微计算机信息》, vol. 22, no. 33, 30 September 2005 (2005-09-30), pages 237 - 239 * |
| 朱凌云: "移动心电监护系统ECG信号的智能检测与分析方法研究", 《中国博士学位论文全文数据库 信息科技辑》, no. 4, 30 April 2004 (2004-04-30), pages 138 - 83 * |
| 赵勇，洪文学: "基于多特征和支持向量机的心律失常分类", 《生物医学工程学杂志》, vol. 28, no. 2, 30 April 2011 (2011-04-30), pages 292 - 295 * |
Cited By (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104382582A (en) * | 2014-11-10 | 2015-03-04 | 哈尔滨医科大学 | Device for classifying dynamic electrocardio data |
| CN104382582B (en) * | 2014-11-10 | 2016-08-31 | 哈尔滨医科大学 | A kind of device that dynamic electrocardiogram (ECG) data is classified |
| CN104398254A (en) * | 2014-11-14 | 2015-03-11 | 中国科学院深圳先进技术研究院 | Electrocardiogram analyzing system, electrocardiogram analyzing equipment and electrocardiogram predication model acquisition equipment |
| CN105769171B (en) * | 2014-12-26 | 2019-01-01 | 中国移动通信集团公司 | The method and device of arrhythmia detection |
| CN105769171A (en) * | 2014-12-26 | 2016-07-20 | 中国移动通信集团公司 | Arrhythmia detection method and device |
| CN104523266A (en) * | 2015-01-07 | 2015-04-22 | 河北大学 | Automatic classification method for electrocardiogram signals |
| CN104523266B (en) * | 2015-01-07 | 2017-04-05 | 河北大学 | A kind of electrocardiosignal automatic classification method |
| US10163528B2 (en) | 2015-11-17 | 2018-12-25 | Anhui Huami Information Technology Co., Ltd. | Determining user-interested information based on wearable device |
| WO2017084546A1 (en) * | 2015-11-17 | 2017-05-26 | 安徽华米信息科技有限公司 | Wearable device-based user interest information determination method, device and wearable device |
| CN105468951B (en) * | 2015-11-17 | 2019-08-06 | 安徽华米信息科技有限公司 | Method and device, the wearable device of identification are carried out by ecg characteristics |
| CN105468951A (en) * | 2015-11-17 | 2016-04-06 | 安徽华米信息科技有限公司 | Method and device for identity recognition through electrocardiographic feature and wearable device |
| US11771359B2 (en) | 2016-01-14 | 2023-10-03 | Biosense Webster (Israel) Ltd. | Region of interest focal source detection using comparisons of R-S wave magnitudes and LATs of RS complexes |
| CN106963481A (en) * | 2016-01-14 | 2017-07-21 | 韦伯斯特生物官能(以色列)有限公司 | Focal source is recognized during atrial fibrillation |
| CN106963481B (en) * | 2016-01-14 | 2021-05-18 | 韦伯斯特生物官能(以色列)有限公司 | Identifying focal sources during atrial fibrillation |
| US11006887B2 (en) | 2016-01-14 | 2021-05-18 | Biosense Webster (Israel) Ltd. | Region of interest focal source detection using comparisons of R-S wave magnitudes and LATs of RS complexes |
| US11806152B2 (en) | 2016-01-14 | 2023-11-07 | Biosense Webster (Israel), Ltd. | Non-overlapping loop-type or spline-type catheter to determine activation source direction and activation source type |
| WO2018184202A1 (en) * | 2017-04-07 | 2018-10-11 | 北京悦琦创通科技有限公司 | Inference engine training method and device, upgrade method, device and storage medium |
| CN106901724A (en) * | 2017-04-26 | 2017-06-30 | 河南省中医院（河南中医药大学第二附属医院） | Electrocardiogram testing device and testing method |
| CN107451417A (en) * | 2017-09-08 | 2017-12-08 | 北京蓬阳丰业医疗设备有限公司 | Dynamic ECG analysis intelligent diagnosis system and method |
| CN111356401A (en) * | 2017-11-29 | 2020-06-30 | 威里利生命科学有限责任公司 | Continuous detection and monitoring of cardiac arrhythmias using wearable sensors and cloud residency analysis |
| CN111356401B (en) * | 2017-11-29 | 2024-03-26 | 威里利生命科学有限责任公司 | Continuous detection and monitoring of cardiac arrhythmias using wearable sensors and cloud-resident analysis |
| CN109222957A (en) * | 2018-08-29 | 2019-01-18 | 西安电子科技大学 | A kind of automation ECG measuring system |
| CN113679347A (en) * | 2021-08-24 | 2021-11-23 | 安徽相王医疗健康股份有限公司 | Diagnosis and treatment effect evaluation method based on data statistics |
| CN113679347B (en) * | 2021-08-24 | 2024-03-19 | 安徽相王医疗健康股份有限公司 | Diagnosis and treatment effect judging method based on data statistics |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103584852A (en) | Personalized electrocardiogram intelligent auxiliary diagnosis device and method | |
| Gao et al. | An effective LSTM recurrent network to detect arrhythmia on imbalanced ECG dataset | |
| Deng et al. | Extracting cardiac dynamics within ECG signal for human identification and cardiovascular diseases classification | |
| JP2022523741A (en) | ECG processing system for depiction and classification | |
| Yan et al. | A restricted Boltzmann machine based two-lead electrocardiography classification | |
| US20230143594A1 (en) | Systems and methods for reduced lead electrocardiogram diagnosis using deep neural networks and rule-based systems | |
| CN113712569B (en) | High-frequency QRS wave group data analysis method and device | |
| CN103038772A (en) | Method of predicting the survivability of a patient | |
| CN107438399A (en) | Angiocarpy deteriorates early warning scoring | |
| CN107072545A (en) | ECG data analysis method and system for quick diagnosis | |
| CN107530019A (en) | Automatic detection/the classification exchanged for the ECG cable of different ECG Lead systems | |
| CN109303559A (en) | A kind of dynamic ECG beat classification method promoting decision tree based on gradient | |
| CN111488793A (en) | Electrocardiosignal classification method and device, electronic equipment and storage medium | |
| Muankid et al. | The Real-time Electrocardiogram Signal Monitoring System in Wireless Sensor Network. | |
| CN105611872A (en) | An apparatus and method for evaluating multichannel ECG signals | |
| Memis et al. | Multimodal classification of obstructive sleep apnea using feature level fusion | |
| CN113080996B (en) | Electrocardiogram analysis method and device based on target detection | |
| US20160242664A1 (en) | An apparatus and method for evaluating multichannel ecg signals | |
| CN109805924A (en) | ECG's data compression method and cardiac arrhythmia detection system based on CNN | |
| US11571161B2 (en) | Systems and methods for electrocardiogram diagnosis using deep neural networks and rule-based systems | |
| CN109044340A (en) | Electrocardiogram (ECG) data classification method, device, electronic equipment and storage medium | |
| CN106974644A (en) | A kind of electrocardiogram R wave detection method and system | |
| TWI688371B (en) | Intelligent device for atrial fibrillation signal pattern acquisition and auxiliary diagnosis | |
| CN206497455U (en) | Ophthalmology image management system | |
| Bahuguna et al. | Normal and Abnormal ECG Signal Classification using Deep Learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20140219 |