CN109276242A - The method and apparatus of electrocardiosignal type identification - Google Patents
The method and apparatus of electrocardiosignal type identification Download PDFInfo
- Publication number
- CN109276242A CN109276242A CN201810870303.7A CN201810870303A CN109276242A CN 109276242 A CN109276242 A CN 109276242A CN 201810870303 A CN201810870303 A CN 201810870303A CN 109276242 A CN109276242 A CN 109276242A
- Authority
- CN
- China
- Prior art keywords
- electrocardiosignal
- type identification
- type
- information
- model
- 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
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/318—Heart-related electrical modalities, e.g. electrocardiography [ECG]
- A61B5/346—Analysis of electrocardiograms
- A61B5/349—Detecting specific parameters of the electrocardiograph cycle
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/318—Heart-related electrical modalities, e.g. electrocardiography [ECG]
- A61B5/346—Analysis of electrocardiograms
- A61B5/349—Detecting specific parameters of the electrocardiograph cycle
- A61B5/35—Detecting specific parameters of the electrocardiograph cycle by template matching
Abstract
The present invention is suitable for data analysis technique field, provides a kind of method and apparatus of electrocardiosignal type identification, comprising: obtains the electrocardiosignal with default source-information;Type identification model is obtained, the type identification model has the type of the electrocardiosignal of the default source-information for identification;The type identification result of the electrocardiosignal is obtained by the type identification model based on the electrocardiosignal;The present invention is by the electrocardiosignal with default source-information by having the type identification model of the type of the electrocardiosignal of default source-information for identification, type identification is obtained as a result, low to electrocardiosignal identification accuracy in the prior art to solve the problems, such as.
Description
Technical field
The invention belongs to data analysis technique field more particularly to a kind of method and apparatus of electrocardiosignal type identification.
Background technique
Heart before shrinking, can have preparatory excitement, and spread to whole body, so that body surface is generated potential difference, use electrocardiograph
These potential differences are recorded, pattern is just electrocardiogram.Electrocardiography is 20th century to set up and be widely used in
One of clinical diagnosis and the important technical achievement of monitoring.With the fast development of computer technology, the standardization of electrocardiogram application
Also it is constantly updating.
American Heart Association's (AHA) joint American Society of Cardiology foundation (ACCF), rhythm of the heart association of the U.S. (HRS) are to the heart
Electrograph standardization and parsing have carried out a series of suggestion and guide, wherein suggest the diagnostic criteria reply race of electrocardiogram it is equal into
Row correction.Traditional electrocardiogram automatic parsing algorithm is mostly that different threshold values is arranged according to current very specific racial difference to sentence
Disconnected, the normal value raised such as V2 lead J point is 0.15mV in Caucasian male's upper limit, negro male 0.20mV, and Caucasian female is
0.10mV, negro male 0.15mV.But this method need to carry out a large amount of electrocardiogram (ECG) data by experienced professional person
Statistical analysis, is difficult with the advantage of big data having counted a large amount of electrocardiogram (ECG) data;Meanwhile only permitting in the selection of feature
Perhaps there is a small amount of parameter, as a result accuracy is not still high.In addition, different medical institutions, hospital and edge such as highlands
The hospital in Haiti area requires criterion to be also not quite similar the analysis of electrocardiogram, complete phase since the medical group faced is different
The medical environment of different zones can not be well adapted to the ecg analysis algorithm of standard, precision of analysis is lower.
Summary of the invention
In view of this, can solve the embodiment of the invention provides a kind of method and apparatus of electrocardiosignal type identification
Prior art electrocardiosignal identifies the low problem of accuracy.
The first aspect of the embodiment of the present invention provides a kind of method of electrocardiosignal type identification, comprising:
Obtain the electrocardiosignal with default source-information;
Type identification model is obtained, there is the type identification model electrocardio of the default source-information to believe for identification
Number type;
The type identification result of the electrocardiosignal is obtained by the type identification model based on the electrocardiosignal.
The second aspect of the embodiment of the present invention provides a kind of method of electrocardiosignal type identification, comprising:
Obtain the source-information of electrocardiosignal and the electrocardiosignal;
According to the source-information, type identification model corresponding with the source-information, the type identification mould are obtained
Type has the type of the electrocardiosignal of the source-information for identification;
The type identification result of the electrocardiosignal is obtained by the type identification model based on the electrocardiosignal.
The third aspect of the embodiment of the present invention provides a kind of method of electrocardiosignal type identification, comprising:
Obtain the type label of sample electrocardiosignal and the sample electrocardiosignal;
Using the sample electrocardiosignal and corresponding type label, training type identification model is trained
Type identification basic model;
The type label for obtaining the target electrocardiosignal and the target electrocardiosignal with identical source information, utilizes institute
Target electrocardiosignal and corresponding type label are stated, the training type identification basic model obtains believing with the source
Corresponding trained type identification differential pattern is ceased, the type identification differential pattern has the source-information for identification
Electrocardiosignal type.
The fourth aspect of the embodiment of the present invention provides a kind of device of electrocardiosignal type identification, comprising:
Acquiring unit, for obtaining the source-information of electrocardiosignal and the electrocardiosignal;
Model unit is determined, for obtaining type identification mould corresponding with the source-information according to the source-information
Type, the type identification model have the type of the electrocardiosignal of the source-information for identification;
Recognition unit, for obtaining the electrocardiosignal by the type identification model based on the electrocardiosignal
Type identification result.
5th aspect of the embodiment of the present invention provides a kind of terminal device, including memory, processor and is stored in
In the memory and the computer program that can run on the processor, when the processor executes the computer program
The step of realizing above-mentioned first aspect or second aspect the method.
6th aspect of the embodiment of the present invention provides a kind of computer readable storage medium, the computer-readable storage
Media storage has computer program, and the computer program realizes above-mentioned first aspect and/or second party when being executed by processor
The step of face and/or third aspect the method.
In the embodiment of the present invention, in the embodiment of the present invention, electrocardiosignal and its source-information are obtained, is believed for separate sources
Breath obtains corresponding electrocardiosignal type identification model, solve do not adapt in the prior art the electrocardiosignal of separate sources into
The problem of row differentiation identifies, takes different type identification models to be identified for the electrocardiosignal of separate sources, improves
The accuracy of electrocardiosignal identification.
Detailed description of the invention
It to describe the technical solutions in the embodiments of the present invention more clearly, below will be to embodiment or description of the prior art
Needed in attached drawing be briefly described, it should be apparent that, the accompanying drawings in the following description is only of the invention some
Embodiment for those of ordinary skill in the art without any creative labor, can also be according to these
Attached drawing obtains other attached drawings.
Fig. 1 is a kind of implementation process schematic diagram of the method for electrocardiosignal type identification provided in an embodiment of the present invention;
Fig. 2 is that one heart of electrocardiogram provided in an embodiment of the present invention claps schematic diagram;
Fig. 3 is the implementation process schematic diagram of the method for another electrocardiosignal type identification provided in an embodiment of the present invention;
Fig. 4 is the implementation process schematic diagram of the method for another electrocardiosignal type identification provided in an embodiment of the present invention;
Fig. 5 is a kind of structural schematic diagram of the device of electrocardiosignal type identification provided in an embodiment of the present invention;
Fig. 6 is a kind of another structural schematic diagram of the device of electrocardiosignal type identification provided in an embodiment of the present invention;
Fig. 7 is a kind of another structural schematic diagram of the device of electrocardiosignal type identification provided in an embodiment of the present invention;
Fig. 8 is the structural schematic diagram of electrocardiograph provided in an embodiment of the present invention.
Specific embodiment
In being described below, for illustration and not for limitation, the tool of such as particular system structure, technology etc is proposed
Body details, to understand thoroughly the embodiment of the present invention.However, it will be clear to one skilled in the art that there is no these specific
The present invention also may be implemented in the other embodiments of details.In other situations, it omits to well-known system, device, electricity
The detailed description of road and method, in case unnecessary details interferes description of the invention.Meanwhile in the description of the present invention,
Term " first " and " second " etc. are only used for distinguishing description, are not understood to indicate or imply relative importance.
In order to illustrate technical solutions according to the invention, the following is a description of specific embodiments.
Fig. 1 shows a kind of implementation process of the method for electrocardiosignal type identification provided in an embodiment of the present invention, the party
Method is suitable for the case where carrying out type identification to electrocardiosignal, is executed by the device of electrocardiosignal type identification.The electrocardiosignal
The device of type identification is generally disposed in terminal device, by software and or hardware realization.Terminal device can be electrocardiograph
Deng.As shown in Figure 1, the method for electrocardiosignal type identification is comprising steps of S101 to S104.
In S101, the source-information of electrocardiosignal and the electrocardiosignal is obtained.
In the embodiment of the present invention, electrocardiosignal is a kind of physiological signal, including human ecg signal and animal body electrocardio
Signal.By the way that the measuring electrode in signal collecting device to be placed on to certain position of organism surface, a system can be acquired out
The electrocardiogram (ECG) data of column, the electrocardiogram (ECG) data have recorded the regular voltage change that parts of body occurs in each cardiac cycle
Situation, then the electrocardiogram (ECG) data for describing the voltage dynamic changes is electrocardiosignal, is shown in the form of ecg wave form
In signal collecting device, wherein the signal collecting device can be electrocardiograph.
A heart in electrocardiogram is clapped, and can have 5 or 6 waves on electrocardiogram, as shown in Fig. 2, be from left to right P wave,
QRS wave, T involve the wave groups such as U wave;Be divided into 7 parts again, i.e., P wave, PR sections, PR interphase, QRS complex, ST sections, T wave, QT interphase,
Wave mode can have inversion because of lead difference.The embodiment of the present invention can be adapted for including standard limb lead and chest V1 lead
Electrocardio measurement pattern is readily applicable to the electrocardio measurement pattern of all kinds of conventional leads systems, and such as 12 standard lead systems, 15 are led
Conjuncted system, 18 lead systems, 9 lead systems etc..
Source-information includes the ethnic information or area information of the electrocardiosignal.Ethnic information reacts collected electrocardio letter
Number human body belonging to race, for example including white people, black race, yellow and brown kind of people.Area information then reacts the collected heart
The affiliated geographic area of the organism of electric signal, for example including plateau, Plain, mountainous region and hills.
The source-information for obtaining electrocardiosignal, to distinguish race or region belonging to electrocardiosignal, thus according to source-information
Determine the type identification model for being used for electrocardiosignal type identification.
In S102, according to the source-information, type identification model corresponding with the source-information, the class are obtained
Type identification model has the type of the electrocardiosignal of the source-information for identification.
In the embodiment of the present invention, after obtaining electrocardiosignal, according to the source-information of electrocardiosignal, acquisition and source-information
Corresponding type identification model, the type identification model have the classification of the electrocardiosignal of the source-information for identification.
The default type identification model for being equipped with corresponding separate sources information, after electrocardiosignal to be identified has been determined, root
According to the source-information of electrocardiosignal, deletes and select type identification model corresponding with source-information.
For example, corresponding plateau, Plain, mountainous region and four kinds of hills area information, there are four types of type identification models, when acquisition
The area information of electrocardiosignal is plateau, then the corresponding type identification model in plateau is obtained, to utilize the type identification model
Identify the classification of electrocardiosignal to be identified.
For another example, four kinds of corresponding white people, black race, yellow and brown kind of people ethnic informations, there are four types of type identification model,
When acquisition electrocardiosignal ethnic information be yellow, then obtain the corresponding type identification model of yellow, thus using should
Type identification model identifies the classification of electrocardiosignal to be identified.
Type identification model has the classification of the electrocardiosignal of the source-information for identification.Wherein, type identification model
It can may be more disaggregated models for two disaggregated models.And more disaggregated models can be more disaggregated models of mutual exclusion, it can also be with
It is multiclass multi-tag model.
As an embodiment of the present invention, type identification model can be two disaggregated models, right respectively when such as output is 0 and 1
The type identification result answered is normal electrocardiosignal and abnormal electrocardiogram signal.
As another embodiment of the present invention, type identification model is more disaggregated models of mutual exclusion, such as the mutual exclusion of electrocardiosignal
Four classification: the atrial arrhythmia that auricular fibrillation, auricular flutter, ventricular arrhythmia, non-room are flutterred/quivered.Auricular fibrillation, atrium
It flutters, the corresponding class probability of atrial arrhythmia that ventricular arrhythmia, non-room are flutterred/quivered is respectively 0.1,0.6,0.15 and
0.15, then the corresponding type identification result of electrocardiosignal is the classification results of maximum probability, i.e. auricular flutter.
As another embodiment of the present invention, type identification model is more classification multi-tag models, such as abnormal electrocardiogram signal
Four classification of multi-tag: arrhythmia cordis, ventricular hypertrophy, myocardial infarction and block.To the type identification result of electrocardiosignal
Only to export tag along sort corresponding with ventricular hypertrophy and block, then the corresponding type identification result of electrocardiosignal is two
It is a, i.e. ventricular hypertrophy and block.
In S103, the electrocardiosignal is pre-processed to obtain pretreated electrocardiosignal;
In the embodiment of the present invention, carrying out pretreatment to electrocardiosignal includes: to be converted to pre- to electrocardiosignal by resampling
It is marked with quasiconfiguaration, and the electrocardiosignal of the preset standard format after conversion is filtered.
Wherein, the data format of electrocardiosignal include but is not limited to data sampling rate, data resolution, lead number and
Electrocardiosignal, is converted to preset standard format by lead sequence etc., i.e., being converted to electrocardiosignal has default sample rate, presets
Resolution ratio, default lead number and default lead sequence etc..Preset standard format depending on specific implementation, the present invention to this not
Do concrete restriction.To electrocardiosignal carry out resampling primarily to uniform data sample rate, data resolution, lead number,
With lead sequence etc..Wherein, by being downsampled to preset a certain frequency after frequency overlapped-resistable filter, for example, by electrocardiosignal from
1000Hz is downsampled to 500Hz, needs to first pass through 250Hz low pass anti-aliasing filter below, then carries out the snap shot of 2:1,
To which the data sampling rate of electrocardiosignal is downsampled to 500Hz from 1000Hz.
Filtering processing includes baseline drift filtering, low-pass filtering or power frequency filtering etc., and filtering processing is for removing electrocardio letter
Noise jamming information in number.In embodiments of the present invention, including but not limited to using average filter, first order IIR filtering device,
Or the modes such as wavelet transformation are filtered the electrocardiosignal of preset standard format.
It should be noted that the sequence of S102 and S103 are adjustable in the embodiment of the present invention, S102 can S103 it
Preceding execution can also execute after S103, or be performed simultaneously with S103.The present invention is successively suitable to the execution of S102 and S103
Sequence is not specifically limited.
In S104, the electrocardio is obtained by the type identification model based on the pretreated electrocardiosignal
The type identification result of signal.
It in embodiments of the present invention, can be based on the pretreated electrocardio letter after being pre-processed to electrocardiosignal
Number the type identification result of the electrocardiosignal is obtained by type identification model.
Optionally, the type identification model can be the deep learning based on machine learning techniques in artificial intelligence
Model, for example, convolutional neural networks model, recursion cycle neural network model and depth confidence network model etc..
As an embodiment of the present invention, it is obtained based on the pretreated electrocardiosignal by the type identification model
To the type identification result of the electrocardiosignal, comprising: the pretreated electrocardiosignal is inputted the type identification mould
Type obtains the type identification result of the electrocardiosignal.
In the present embodiment, directly by pretreated electrocardiosignal, i.e., the electricity that is composed of each sampling instant voltage value
Sequence is pressed, as the input vector of type identification model, input type identification model obtains the type identification of the electrocardiosignal
As a result.
It is described to be based on the pretreated electrocardiosignal as another embodiment of the present invention, pass through the type identification
Model obtains the type identification result of the electrocardiosignal, comprising: extracts the feature ginseng of the pretreated electrocardiosignal
Number generates feature vector according to the characteristic parameter, described eigenvector is inputted the type identification model, obtains the heart
The type identification result of electric signal.
Specifically, the characteristic parameter of the pretreated electrocardiosignal is extracted, comprising: the pretreated heart of detection
QRS wave in electric signal extracts the electrocardio rhythm and pace of moving things of the pretreated electrocardiosignal based on the testing result of the QRS wave
Information, and electrocardio measurement parameter is calculated based on the electrocardio rhythm and pace of moving things information, the electrocardio rhythm and pace of moving things information and/or the electrocardio are surveyed
Measure characteristic parameter of the parameter as the pretreated electrocardiosignal.
Wherein, QRS wave refers to the maximum wave group of amplitude in normal ECG, is able to reflect the overall process of sequences of ventricular depolarization.Just
Normal sequences of ventricular depolarization starts from the middle part of interventricular septum, from left to right direction depolarization, therefore one small downward Q wave is first presented in QRS complex.
It, can be from electrocardiosignal using the QRS detection algorithm such as calculus of finite differences, threshold detection method, template matching method, Wavelet Transform
In detect position where each QRS complex, to obtain each QRS complex in electrocardiosignal.
In embodiments of the present invention, classical Pan-Tompkins method can be selected to carry out the detection of QRS wave, this method packet
It includes bandpass filtering, nonlinear transformation and rule and judges three steps, the bandpass filter is by high-pass filter and low-pass filtering
Device cascades, and enhances the signal component of 5~12Hz frequency range where QRS wave main energetic;Then signal is carried out point-by-point micro-
Point, square and integral operation, the signal after obtaining nonlinear transformation;If it is detected that the peak value of signal is greater than preset threshold, recognize
To detect a QRS wave.
After carrying out QRS wave detection, the electrocardio rhythm and pace of moving things information of pretreated electrocardiosignal is extracted.Electrocardio rhythm and pace of moving things information includes
The position of each heart is clapped in electrocardiosignal P wave, QRS wave and T wave.
Electrocardio measurement parameter is calculated based on electrocardio rhythm and pace of moving things information.Electrocardio measurement parameter includes: each wavelet amplitude or time limit, respectively
Wavelet interphase.Each wavelet amplitude or time limit include but is not limited to: amplitude limit value, T wave time limit when each wavelet of P wave time limit amplitude, QRS wave
Amplitude and ST sections of amplitudes;Each wavelet interphase includes but is not limited to: PR interphase, QT interphase, RR interphase and PP interphase.
Based on the electrocardio rhythm and pace of moving things information and/or the electrocardio measurement parameter as the pretreated electrocardiosignal
Characteristic parameter.Then selected section or whole characteristic parameter constitutive characteristic vectors are obtained as the input vector of classification identification model
To electrocardiosignal type identification result.Wherein, selected section or whole characteristic parameter constitutive characteristic vectors, based in training type
When identification model, which characteristic parameter constitutive characteristic vector is taken to be trained, then accordingly, using trained type
When identification model, then corresponding feature vector input type identification model is generated with output category result.
In addition, reducing the influence of individual difference in order to make type identification model be suitable for Different Individual, needing to extraction
The characteristic parameter be standardized, and after the standardization for completing the characteristic parameter, by the characteristic parameter structure
At feature vector.Wherein, the standardization is a kind of dimensionless processing means, for by the absolute value of the characteristic parameter numerical value
Become the relationship of certain relative value, simplifies the Generalization Capability for improving the type identification model while calculating.
It is pretreated based on what is extracted on the basis of both of the aforesaid embodiment as another embodiment of the present invention
The characteristic parameter and the pretreated electrocardiosignal are combined into input vector by the characteristic parameter of the electrocardiosignal,
The type identification model is inputted, the type identification result of the electrocardiosignal is obtained.
In the present embodiment, characteristic parameter and electrocardiosignal are combined into input vector, directly improve input vector
Dimension, so that input vector carries richer characteristic information, so as to obtain more accurate recognition result.
Optionally, in order to make type identification model be suitable for Different Individual, reduce the influence of individual difference, need to mentioning
The characteristic parameter and electrocardiosignal taken is standardized, and in the standard for completing the characteristic parameter and electrocardiosignal
After change, constitutive characteristic vector the characteristic parameter and electrocardiosignal is combined.
After generating described eigenvector, described eigenvector need to be only input to the trained type identification mould
The type identification result of electrocardiosignal can be obtained in type.
In the embodiment of the present invention, electrocardiosignal and its source-information are obtained, for the corresponding heart of separate sources acquisition of information
Electrical signal types identification model improves the accuracy of electrocardiosignal identification.
On the basis of the above embodiments, it should be noted that if not being too high feelings to the required precision of recognition result
Under condition, it can also omit and pretreated step is carried out to electrocardiosignal, at this point, directly carrying out type to collected electrocardiosignal
Identification, without carrying out type identification by type identification model again after pre-processing to electrocardiosignal.Electrocardiosignal at this time
It does not include that pretreated step is carried out to electrocardiosignal in kind identification method, other steps are similar to the above embodiments, herein
It repeats no more.
As shown in figure 3, in addition the embodiment of the present invention also provides a kind of method of electrocardiosignal type identification, the identification side
Method is available for the trained type identification model of identification electrocardiosignal type, and this method is by electrocardiosignal type identification
Device execute.The device of the electrocardiosignal type identification is generally disposed in server, can be implemented by software and/or hardware.Such as
Shown in Fig. 3, the method for the electrocardiosignal type identification is comprising steps of S301 to S304.
In S301, the type label of sample electrocardiosignal and the sample electrocardiosignal is obtained.
In embodiments of the present invention, sample electrocardiosignal forms basic database.Electrocardiogram (ECG) data in the basic database
It can be and collected by the electrocardiograms such as electrocardiograph, electrocardio network or electrocardio cloud platform acquisition equipment, it can also be by all kinds of
Disclosed ECG data composition.Whole sample electrocardiogram (ECG) datas in basic database are marked by clinician or expert
Note, thus obtains type label corresponding with sample electrocardiosignal.For example, when the corresponding classification of sample electrocardiosignal is normal
When electrocardiosignal, which is labeled as 1;When the corresponding classification of sample electrocardiosignal is different
When normal electrocardiosignal, which is labeled as 0.
It should be noted that the ECG data of the basic database derived from multiracial or multizone distribution compared with
Wide ECG data collects.The type label of sample electrocardiosignal is the corresponding label of type of the electrocardiosignal.
Sample electrocardiogram (ECG) data covers electrocardiosignal all types to be sorted.For example, to train electrocardiosignal for identification is to belong to
In two Classification and Identification models of normal electrocardiosignal and abnormal electrocardiogram signal, then the type label of sample electrocardiosignal includes normal
Electrocardiosignal label and abnormal electrocardiogram signal label.I.e. sample electrocardiogram (ECG) data includes normal electrocardiosignal and abnormal electrocardiogram signal.
In S302, the sample electrocardiosignal is pre-processed to obtain pretreated sample electrocardiosignal.
Wherein, it carries out pretreated detailed process to the sample electrocardiosignal and is referred to abovementioned steps S103 to retouch
It states, details are not described herein again.
In S303, the pretreated sample electrocardiosignal and corresponding type label, training type are utilized
Identification model obtains trained type identification basic model;
After pre-processing to the sample electrocardiosignal, according to corresponding type label, training class can be passed through
Type identification model obtains trained type identification basic model.
Specifically, S303 may include: to generate input vector according to the pretreated sample electrocardiosignal, will be described
Input vector inputs the type identification model, is exported as a result, according to the output result and the pretreated sample
The corresponding tag types of this electrocardiosignal adjust input vector, model structure or the hyper parameter of the type identification model, until
The type identification model meets preset condition, will meet the type identification model of preset condition as type identification basis mould
Type.
Wherein, according to the description to embodiment illustrated in fig. 1, accordingly, for training the input vector of type identification model
There are three types of situations, and details are not described herein again.
In the training process, use pattern identification model carries out type identification to above-mentioned sample electrocardiosignal, according to identification
As a result the input vector or at least one of model structure or hyper parameter for adjusting type identification model, until class adjusted
Type identification model can identify the type of all sample electrocardiosignals, or detect that the accuracy rate of type identification is greater than preset value
Deng then using the type identification model adjusted as trained type identification basic model.Wherein, adjustment model structure can
To include changing model classification, that is, deep learning model is changed, such as convolutional neural networks model, recursion cycle neural network mould
Type and depth confidence network model etc..The hyper parameter of type identification model may include learning rate, the number of iterations, every layer
The number etc. of neuron.The input vector of type identification model may include characteristic parameter and dimension etc. that input vector includes.
In S304, the type mark of the target electrocardiosignal and the target electrocardiosignal with identical source information is obtained
Label, using the target electrocardiosignal and corresponding type label, the training type identification basic model is obtained and institute
The corresponding trained type identification differential pattern of source-information is stated, the type identification differential pattern has described for identification
The type of the electrocardiosignal of source-information.
Since basic database is the electrocardiogram (ECG) data set that multiple sources are collected, which may includes a certain
Particular source, such as yellow, electrocardiogram, may not also include this particular source electrocardiogram;It is instructed by basic database
The type identification model got can make correct classification to the electrocardiogram of general population, but work as the electrocardio of particular source
When figure feature and general population's electrocardiogram are not quite identical, the classification results of type identification basic model may mistake;For
The classification error to the electrocardiogram of particular source is reduced, introduces transfer learning thought, i.e. step S304.
As an embodiment of the present invention, obtaining in step S304 has the target electrocardiosignal of identical source information and described
The type label of target electrocardiosignal, comprising: filter out the sample with identical source information in the sample electrocardiosignal
Electrocardiosignal, and using the sample electrocardiosignal with identical source information as target electrocardiosignal;And obtain the target heart
The corresponding type label of electric signal.
In this embodiment, if having the ECG data of certain particular source in basic database, from basic database
The partial data is screened, as target area database, continuation is instructed on the basis of type identification basic model
Practice, this process increases the weight of the electrocardiogram (ECG) data of particular source, improves the electrocardio of the organism for particular source
The classification accuracy of signal.
As another embodiment of the present invention, the target electrocardiosignal with identical source information and institute are obtained in step S304
State the type label of target electrocardiosignal, comprising: obtain the target electrocardiosignal with identical source information, and obtain the mesh
Mark the corresponding type label of electrocardiosignal.
In this embodiment, if not having the electrocardiogram (ECG) data of certain particular source in basic database, it is specific next to collect this
The electrocardiogram (ECG) data in source continues the electrocardiogram (ECG) data of collection on the basis of type identification basic model as target area database
Upper training, in the training process, basic model can further learn to extract the ecg characteristics of this particular source, improve model
To the classification accuracy of the electrocardiogram in the source.
It should be noted that the training type identification basic model obtains the process and training of type identification differential pattern
The process that type identification model obtains type identification basic model is similar, adjusts type identification basis mould in the training process
The input vector and hyper parameter of type, until obtaining the type identification differential pattern for meeting preset condition.
On the basis of the above embodiments, it should be noted that if not being too high feelings to the required precision of recognition result
It, can also be without carrying out pretreated step to electrocardiosignal, at this point, collected electrocardiosignal is directly carried out class under condition
The training of type identification model.It does not at this time include that pretreated step is carried out to electrocardiosignal in the kind identification method of electrocardiosignal
Suddenly, other steps are similar to the above embodiments, and details are not described herein again.
The type identification differential pattern of trained multiple information of separate sources for identification is pushed to electrocardio by server
Figure machine, or be pre-stored before electrocardiograph factory by research staff, thus in the electrocardiosignal acquired to electrocardiograph
When carrying out type identification, obtains type identification differential pattern corresponding with the source-information of electrocardiosignal and carry out type identification.It needs
It is noted that the electrocardiosignal of electrocardiograph acquisition can also feed back to server, to increase the electrocardio number of basic database
According in advanced optimizing for server end execution type identification differential pattern, the type identification that server will further optimize again
Differential pattern is pushed to electrocardiograph with the type identification differential pattern before replacing.
In other embodiments, trained type identification differential pattern is pushed to certain region or certain ethnic group collection by server
In electrocardiograph, or be pre-stored before the electrocardiograph factory that certain region or certain ethnic group are concentrated by research staff, from
And when the electrocardiosignal acquired to electrocardiograph carries out type identification, directly acquire the electrocardio letter with default source-information
Number, and obtain type identification differential pattern and type identification directly is carried out to the electrocardiosignal with default source-information.It needs
Bright, the electrocardiosignal of the electrocardiograph acquisition of certain region or certain ethnic group concentration can also feed back to server, to increase base
The electrocardiogram (ECG) data of plinth database and/or target area database executes the further of type identification differential pattern in server end
Optimization, the type identification differential pattern that server will further optimize again are pushed to corresponding electrocardiograph with the class before replacing
Type Recognition Different model.Alternatively, the electrocardiosignal for the electrocardiograph acquisition that certain region or certain ethnic group are concentrated feeds back to server, group
At the electrocardiogram (ECG) data of the target area database during server retraining, type identification basic model is executed in server end
It is further training to obtain type identification differential pattern, this type identification differential pattern is pushed to electrocardiogram again by server
Machine.In this case, electrocardiograph only has a kind of type identification differential pattern of corresponding default source-information, works as electrocardiograph
After collecting the electrocardiosignal with default source-information, the electrocardiosignal for identification with the default source-information is got
Differential pattern carry out type identification.
Based on this, the present invention also provides the methods of another electrocardiosignal type identification.Fig. 4 shows implementation of the present invention
The implementation process of the method for another electrocardiosignal type identification that example provides, this method are suitable for carrying out type to electrocardiosignal
The case where identification, is executed by the device of electrocardiosignal type identification.The device of the electrocardiosignal type identification is generally disposed in end
End equipment, by software and or hardware realization.Terminal device can be electrocardiograph etc..As shown in figure 4, electrocardiosignal type is known
Method for distinguishing is comprising steps of S401 to S403.Place is not described in detail in this method embodiment please participate in embodiment described in Fig. 1.
In S401, the electrocardiosignal with default source-information is obtained.
Wherein, since electrocardiograph is certain region or for the electrocardiograph of certain ethnic group, electrocardiograph acquisition
Electrocardiosignal is the electrocardiosignal with default source-information.Default source-information includes any of area information or ethnic group information
Kind.For example, electrocardiograph is the electrocardiograph of highlands, then the electrocardiosignal from highlands is obtained.
In S402, type identification model is obtained, the type identification model has the default source letter for identification
The type of the electrocardiosignal of breath.
Wherein, electrocardiograph presets the type identification mould that identification has the type for the electrocardiosignal for presetting source-information
Type.For example, electrocardiograph is the electrocardiograph of highlands, then the type identification of the electrocardiosignal of identification highlands is preset
Model.
In S403, it is based on the electrocardiosignal, by the type identification model, the type for obtaining the electrocardiosignal is known
Other result.
Optionally, on the basis of the above embodiments, further include that pretreated step is carried out to electrocardiosignal, refer to pair
The description of embodiment illustrated in fig. 1, details are not described herein again.
The validity of the method for the present invention introducing transfer learning thought is verified with specific test below.
Basic database data source in test passes through the acquisition of the platforms such as electrocardiograph, electrocardio network in more ground hospital
Electrocardiogram, about 150,000.The electrocardiogram that target area database source is acquired in somewhere by electrocardio network, about 50,000.Institute is intentionally
Electric data pass through clinician or expert's mark, and 80% is used as training set in the database of target area, and 10% collects as verifying,
10% is used as test set.
The type identification differential pattern that table 1 is given type identification basic model and obtained using retraining after transfer learning
To the accuracy of target area ECG data analysis.
Table 1: type identification basic model (identifying A in corresponding table) and type identification differential pattern (identifying B in corresponding table)
To the accuracy of somewhere electrocardiogram classification
Recall rate | Specificity | Accuracy rate | Accuracy | F score | |
A | 90.73% | 90.37% | 93.87% | 90.59% | 92.27% |
B | 92.36% | 93.99% | 96.54% | 92.94% | 94.40% |
It can be seen from Table 1 that type identification differential pattern of the invention is significantly excellent to the type identification of electrocardiosignal
In type identification basic model, precision of analysis is significantly improved.
As shown in figure 5, being a kind of device 500 of electrocardiosignal type identification provided in an embodiment of the present invention, comprising: obtain
Unit 501 determines model unit 502, pretreatment unit 503 and recognition unit 504.
Acquiring unit 501, for obtaining the source-information of electrocardiosignal and the electrocardiosignal;
Model unit 502 is determined, for obtaining type identification corresponding with the source-information according to the source-information
Model, the type identification model have the type of the electrocardiosignal of the source-information for identification;
Pretreatment unit 503 obtains pretreated electrocardiosignal for being pre-processed to the electrocardiosignal;
Recognition unit 504, for being obtained based on the pretreated electrocardiosignal by the type identification model
The type identification result of the electrocardiosignal.
It should be noted that a kind of realization process of the device of electrocardiosignal type identification provided in this embodiment can join
A kind of realization process of the method for the electrocardiosignal type identification provided as shown in figure 1 is examined, details are not described herein.
As shown in fig. 6, being a kind of device 600 of electrocardiosignal type identification provided in an embodiment of the present invention, comprising: first
Acquiring unit 601, second acquisition unit 602 and recognition unit 603.
First acquisition unit 601, for obtaining the electrocardiosignal with default source-information;
Second acquisition unit 602, for obtaining type identification model, the type identification model has described for identification
The type of the electrocardiosignal of default source-information;
Recognition unit 603, by the type identification model, obtains the electrocardio letter for being based on the electrocardiosignal
Number type identification result.
It should be noted that a kind of realization process of the device of electrocardiosignal type identification provided in this embodiment can join
The realization process of the method such as a kind of electrocardiosignal type identification provided in Fig. 4 is examined, details are not described herein.
As shown in fig. 7, for the device 700 of another electrocardiosignal type identification provided in an embodiment of the present invention, comprising: obtain
Take unit 701, pretreatment unit 702, the first training unit 703 and the second training unit 704.
Acquiring unit 701, for obtaining the type label of sample electrocardiosignal and the sample electrocardiosignal;
Pretreatment unit 702 obtains pretreated sample electrocardio for being pre-processed to the sample electrocardiosignal
Signal;
First training unit 703, for utilizing the pretreated sample electrocardiosignal and corresponding type mark
Label, training type identification model, obtain trained type identification basic model;
Second training unit 704, for obtaining target electrocardiosignal and the target electrocardio with identical source information
The type label of signal, using the target electrocardiosignal and corresponding type label, the training type identification basis
Model, obtains trained type identification differential pattern corresponding with the source-information, and the type identification differential pattern is used
There is the type of the electrocardiosignal of the source-information in identification.
It should be noted that a kind of realization process of the device of electrocardiosignal type identification provided in this embodiment can join
The realization process of the method such as a kind of electrocardiosignal type identification provided in Fig. 3 is examined, details are not described herein.
It should be understood that the size of the serial number of each step is not meant that the order of the execution order in above-described embodiment, each process
Execution sequence should be determined by its function and internal logic, the implementation process without coping with the embodiment of the present invention constitutes any limit
It is fixed.
Fig. 8 is the schematic diagram for the electrocardiograph that one embodiment of the invention provides.As shown in figure 8, the electrocardiogram of the embodiment
Machine 8 includes: processor 80, memory 81 and is stored in the meter that can be run in the memory 81 and on the processor 80
Calculation machine program 82, such as the program of electrocardiosignal type identification.The realization when processor 80 executes the computer program 82
Step in the embodiment of the method for above-mentioned electrocardiosignal type identification, such as step S101 to S104 shown in FIG. 1.Alternatively, institute
The function that each module/unit in above-mentioned each Installation practice is realized when processor 80 executes the computer program 82 is stated, for example,
The function of unit 501 to 504 shown in Fig. 5.
Illustratively, the computer program 82 can be divided into one or more module/units, it is one or
Multiple module/units are stored in the memory 81, and are executed by the processor 80, to complete the present invention.Described one
A or multiple module/units can be the series of computation machine program instruction section that can complete specific function, which is used for
Implementation procedure of the computer program 82 in the electrocardiograph 80 is described.For example, the computer program 82 can be by
It is divided into acquiring unit, determines model unit, pretreatment unit and recognition unit (unit in virtual bench), each unit is specific
Function is as follows:
Acquiring unit, for obtaining the source-information of electrocardiosignal and the electrocardiosignal;
Model unit is determined, for obtaining type identification mould corresponding with the source-information according to the source-information
Type, the type identification model have the type of the electrocardiosignal of the source-information for identification;
Pretreatment unit obtains pretreated electrocardiosignal for being pre-processed to the electrocardiosignal;
Recognition unit, by the type identification model, obtains described for being based on the pretreated electrocardiosignal
The type identification result of electrocardiosignal.
The electrocardiograph may include, but be not limited only to, processor 80, memory 81.Those skilled in the art can manage
Solution, Fig. 8 is only the example of electrocardiograph 80, does not constitute the restriction to electrocardiograph 80, may include than illustrate it is more or
Less component perhaps combines certain components or different components, such as the electrocardiograph can also include input and output
Equipment, network access equipment, bus etc..
Alleged processor 80 can be central processing unit (Central Processing Unit, CPU), can also be
Other general processors, digital signal processor (Digital Signal Processor, DSP), specific integrated circuit
(Application Specific Integrated Circuit, ASIC), field programmable gate array (Field-
Programmable Gate Array, FPGA) either other programmable logic device, discrete gate or transistor logic,
Discrete hardware components etc..General processor can be microprocessor or the processor is also possible to any conventional processor
Deng.
The memory 81 can be the internal storage unit of the electrocardiograph 80, for example, electrocardiograph 80 hard disk or
Memory.The memory 81 is also possible to the External memory equipment of the electrocardiograph 80, such as matches on the electrocardiograph 80
Standby plug-in type hard disk, intelligent memory card (Smart Media Card, SMC), secure digital (Secure Digital, SD)
Card, flash card (Flash Card) etc..Further, the memory 81 can also both include the inside of the electrocardiograph 80
Storage unit also includes External memory equipment.The memory 81 is for storing the computer program and the electrocardiograph
Other required programs and data.The memory 81 can be also used for temporarily storing the number that has exported or will export
According to.
It is apparent to those skilled in the art that for convenience of description and succinctly, only with above-mentioned each function
Can unit, module division progress for example, in practical application, can according to need and by above-mentioned function distribution by different
Functional unit, module are completed, i.e., the internal structure of described device is divided into different functional unit or module, more than completing
The all or part of function of description.Each functional unit in embodiment, module can integrate in one processing unit, can also
To be that each unit physically exists alone, can also be integrated in one unit with two or more units, it is above-mentioned integrated
Unit both can take the form of hardware realization, can also realize in the form of software functional units.In addition, each function list
Member, the specific name of module are also only for convenience of distinguishing each other, the protection scope being not intended to limit this application.Above system
The specific work process of middle unit, module, can refer to corresponding processes in the foregoing method embodiment, and details are not described herein.
The embodiment of the present invention also provides a kind of server, including memory, processor and storage are in the memory
And the computer program that can be run on the processor, the processor realize above-mentioned Fig. 3 when executing the computer program
The step of method of the electrocardiosignal type identification.Alternatively, being realized when the processor execution computer program above-mentioned
The function of each module/unit in each Installation practice, for example, the function of module 701 to 704 shown in Fig. 7.
In the above-described embodiments, it all emphasizes particularly on different fields to the description of each embodiment, is not described in detail or remembers in some embodiment
The part of load may refer to the associated description of other embodiments.
Those of ordinary skill in the art may be aware that list described in conjunction with the examples disclosed in the embodiments of the present disclosure
Member and algorithm steps can be realized with the combination of electronic hardware or computer software and electronic hardware.These functions are actually
It is implemented in hardware or software, the specific application and design constraint depending on technical solution.Professional technician
Each specific application can be used different methods to achieve the described function, but this realization is it is not considered that exceed
The scope of the present invention.
In embodiment provided by the present invention, it should be understood that disclosed device/terminal device and method, it can be with
It realizes by another way.For example, device described above/terminal device embodiment is only schematical, for example, institute
The division of module or unit is stated, only a kind of logical function partition, there may be another division manner in actual implementation, such as
Multiple units or components can be combined or can be integrated into another system, or some features can be ignored or not executed.Separately
A bit, shown or discussed mutual coupling or direct-coupling or communication connection can be through some interfaces, device
Or the INDIRECT COUPLING or communication connection of unit, it can be electrical property, mechanical or other forms.
The unit as illustrated by the separation member may or may not be physically separated, aobvious as unit
The component shown may or may not be physical unit, it can and it is in one place, or may be distributed over multiple
In network unit.It can select some or all of unit therein according to the actual needs to realize the mesh of this embodiment scheme
's.
It, can also be in addition, the functional units in various embodiments of the present invention may be integrated into one processing unit
It is that each unit physically exists alone, can also be integrated in one unit with two or more units.Above-mentioned integrated list
Member both can take the form of hardware realization, can also realize in the form of software functional units.
If the integrated module/unit be realized in the form of SFU software functional unit and as independent product sale or
In use, can store in a computer readable storage medium.Based on this understanding, the present invention realizes above-mentioned implementation
All or part of the process in example method, can also instruct relevant hardware to complete, the meter by computer program
Calculation machine program can be stored in a computer readable storage medium, the computer program when being executed by processor, it can be achieved that on
The step of stating each embodiment of the method.Wherein, the computer program includes computer program code, the computer program generation
Code can be source code form, object identification code form, executable file or certain intermediate forms etc..The computer-readable medium
It may include: any entity or device, recording medium, USB flash disk, mobile hard disk, magnetic that can carry the computer program code
Dish, CD, computer storage, read-only memory (Read-Only Memory, ROM), random access memory (Random
Access Memory, RAM), electric carrier signal, telecommunication signal and software distribution medium etc..It should be noted that the meter
The content that calculation machine readable medium includes can carry out increase and decrease appropriate according to the requirement made laws in jurisdiction with patent practice,
It such as does not include electric carrier signal and telecommunications according to legislation and patent practice, computer-readable medium in certain jurisdictions
Signal.
Embodiment described above is merely illustrative of the technical solution of the present invention, rather than its limitations;Although referring to aforementioned reality
Applying example, invention is explained in detail, those skilled in the art should understand that: it still can be to aforementioned each
Technical solution documented by embodiment is modified or equivalent replacement of some of the technical features;And these are modified
Or replacement, the spirit and scope for technical solution of various embodiments of the present invention that it does not separate the essence of the corresponding technical solution should all
It is included within protection scope of the present invention.
Claims (10)
1. a kind of method of electrocardiosignal type identification characterized by comprising
Obtain the electrocardiosignal with default source-information;
Type identification model is obtained, the type identification model has the electrocardiosignal of the default source-information for identification
Type;
The type identification result of the electrocardiosignal is obtained by the type identification model based on the electrocardiosignal.
2. a kind of method of electrocardiosignal type identification characterized by comprising
Obtain the source-information of electrocardiosignal and the electrocardiosignal;
According to the source-information, type identification model corresponding with the source-information is obtained, the type identification model is used
There is the type of the electrocardiosignal of the source-information in identification;
The type identification result of the electrocardiosignal is obtained by the type identification model based on the electrocardiosignal.
3. method according to claim 1 or 2, which is characterized in that it is described to be based on the electrocardiosignal, known by the type
Other model obtains the type identification result of the electrocardiosignal, comprising:
The characteristic parameter for extracting the electrocardiosignal generates feature vector according to the characteristic parameter, and described eigenvector is defeated
Enter the type identification model, obtains the type identification result of the electrocardiosignal;Or
The characteristic parameter and the electrocardiosignal are combined into input vector by the characteristic parameter for extracting the electrocardiosignal, defeated
Enter the type identification model, obtains the type identification result of the electrocardiosignal;Or
The electrocardiosignal is inputted into the type identification model, obtains the type identification result of the electrocardiosignal.
4. method according to claim 1 or 2, which is characterized in that type identification model includes deep learning model.
5. a kind of method of electrocardiosignal type identification characterized by comprising
Obtain the type label of sample electrocardiosignal and the sample electrocardiosignal;
Using the sample electrocardiosignal and corresponding type label, training type identification model obtains trained class
Type identifies basic model;
The type label for obtaining the target electrocardiosignal and the target electrocardiosignal with identical source information, utilizes the mesh
Electrocardiosignal and corresponding type label are marked, the training type identification basic model obtains and the source-information pair
The trained type identification differential pattern answered, the type identification differential pattern have the heart of the source-information for identification
The type of electric signal.
6. method as claimed in claim 5, which is characterized in that utilize the sample electrocardiosignal and corresponding type mark
Label, training type identification model, obtain trained type identification basic model, comprising:
Input vector is generated according to the sample electrocardiosignal, the input vector is inputted into the type identification model, is obtained
Output is as a result, according to the output result and the corresponding tag types of the pretreated sample electrocardiosignal, described in adjustment
Input vector, model structure or the hyper parameter of type identification model will expire until the type identification model meets preset condition
The type identification model of sufficient preset condition is as type identification basic model.
7. such as method described in claim 5 or 6, which is characterized in that obtain the target electrocardiosignal with identical source information
With the type label of the target electrocardiosignal, comprising:
The sample electrocardiosignal with identical source information is filtered out in the sample electrocardiosignal, and there will be identical source
The sample electrocardiosignal of information is as target electrocardiosignal;Obtain the corresponding type label of the target electrocardiosignal.
8. a kind of device of electrocardiosignal type identification characterized by comprising
Acquiring unit, for obtaining the source-information of electrocardiosignal and the electrocardiosignal;
Model unit is determined, for obtaining type identification model corresponding with the source-information, institute according to the source-information
State type identification model for identification and have the type of the electrocardiosignal of the source-information;
Recognition unit, for obtaining the type of the electrocardiosignal by the type identification model based on the electrocardiosignal
Recognition result.
9. a kind of terminal device, including memory, processor and storage are in the memory and can be on the processor
The computer program of operation, which is characterized in that the processor realizes such as claim 1 to 7 when executing the computer program
The step of any one the method.
10. a kind of computer readable storage medium, the computer-readable recording medium storage has computer program, and feature exists
In when the computer program is executed by processor the step of any one of such as claim 1 to 7 of realization the method.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810870303.7A CN109276242A (en) | 2018-08-02 | 2018-08-02 | The method and apparatus of electrocardiosignal type identification |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810870303.7A CN109276242A (en) | 2018-08-02 | 2018-08-02 | The method and apparatus of electrocardiosignal type identification |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109276242A true CN109276242A (en) | 2019-01-29 |
Family
ID=65182558
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810870303.7A Pending CN109276242A (en) | 2018-08-02 | 2018-08-02 | The method and apparatus of electrocardiosignal type identification |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109276242A (en) |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110037680A (en) * | 2019-03-12 | 2019-07-23 | 平安科技(深圳)有限公司 | The recognition methods of electrocardiogram and device, storage medium, computer equipment |
CN110693491A (en) * | 2019-11-13 | 2020-01-17 | 歌尔科技有限公司 | Electrocardiosignal processing method and device, wearable device and readable storage medium |
CN110693486A (en) * | 2019-09-27 | 2020-01-17 | 武汉中旗生物医疗电子有限公司 | Electrocardiogram abnormity labeling method and device |
CN110720894A (en) * | 2019-11-08 | 2020-01-24 | 广州视源电子科技股份有限公司 | Atrial flutter detection method, device, equipment and storage medium |
CN110897630A (en) * | 2019-06-27 | 2020-03-24 | 上海数创医疗科技有限公司 | Ventricular tachycardia and heart rhythm identification method and device based on transfer learning |
CN111134662A (en) * | 2020-02-17 | 2020-05-12 | 武汉大学 | Electrocardio abnormal signal identification method and device based on transfer learning and confidence degree selection |
CN111265194A (en) * | 2019-12-11 | 2020-06-12 | 广州视源电子科技股份有限公司 | Ventricular hypertrophy detection method and device, storage medium and processor |
CN111466911A (en) * | 2020-04-13 | 2020-07-31 | 徐州工业职业技术学院 | Electronic signal processing system and method |
CN111513702A (en) * | 2020-04-30 | 2020-08-11 | 何中煜 | Method for detecting electrocardiographic data and storage medium |
WO2021037102A1 (en) * | 2019-08-27 | 2021-03-04 | 中科麦迪人工智能研究院(苏州)有限公司 | Electrocardiogram analysis method and apparatus based on picture and heartbeat information, and device and medium |
CN114305446A (en) * | 2021-12-25 | 2022-04-12 | 肇庆星网医疗科技有限公司 | Atrial fibrillation detection method and system based on artificial intelligence |
Citations (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103413063A (en) * | 2013-08-30 | 2013-11-27 | 北京市医疗器械检验所 | Medical diagnostic equipment accuracy assessment method and device |
CN104783782A (en) * | 2015-04-13 | 2015-07-22 | 深圳市飞马与星月科技研究有限公司 | Automatic detection method and device for electrocardiosignals |
CN104970789A (en) * | 2014-04-04 | 2015-10-14 | 中国科学院苏州纳米技术与纳米仿生研究所 | Electrocardiogram classification method and system |
CN105411578A (en) * | 2016-01-02 | 2016-03-23 | 无锡桑尼安科技有限公司 | Automatic patient state detection method |
CN105640521A (en) * | 2016-01-02 | 2016-06-08 | 无锡桑尼安科技有限公司 | Intelligent physiological parameter detection method |
CN106073765A (en) * | 2016-05-27 | 2016-11-09 | 成都信汇聚源科技有限公司 | Die suddenly in a kind of miniature holter devices the implementation method of risk profile |
CN106108889A (en) * | 2016-07-20 | 2016-11-16 | 杨平 | Electrocardiogram classification method based on degree of depth learning algorithm |
US20160346512A1 (en) * | 2013-03-14 | 2016-12-01 | Angiodynamics, Inc. | Systems and methods for catheter tip placement using ecg |
CN106214123A (en) * | 2016-07-20 | 2016-12-14 | 杨平 | A kind of electrocardiogram compressive classification method based on degree of depth learning algorithm |
CN106343976A (en) * | 2016-09-14 | 2017-01-25 | 京东方科技集团股份有限公司 | Method and device for establishing blood pressure model as well as method and device for determining blood pressure |
WO2017048684A1 (en) * | 2015-09-18 | 2017-03-23 | Spangler Scientific Llc | Non-invasive determination of risk for sudden cardiac death |
CN107951485A (en) * | 2017-11-27 | 2018-04-24 | 乐普(北京)医疗器械股份有限公司 | Ambulatory ECG analysis method and apparatus based on artificial intelligence self study |
CN108095708A (en) * | 2018-01-19 | 2018-06-01 | 动析智能科技有限公司 | A kind of physiology monitoring and analysis method, system based on mixing sensing |
-
2018
- 2018-08-02 CN CN201810870303.7A patent/CN109276242A/en active Pending
Patent Citations (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20160346512A1 (en) * | 2013-03-14 | 2016-12-01 | Angiodynamics, Inc. | Systems and methods for catheter tip placement using ecg |
CN103413063A (en) * | 2013-08-30 | 2013-11-27 | 北京市医疗器械检验所 | Medical diagnostic equipment accuracy assessment method and device |
CN104970789A (en) * | 2014-04-04 | 2015-10-14 | 中国科学院苏州纳米技术与纳米仿生研究所 | Electrocardiogram classification method and system |
CN104783782A (en) * | 2015-04-13 | 2015-07-22 | 深圳市飞马与星月科技研究有限公司 | Automatic detection method and device for electrocardiosignals |
WO2017048684A1 (en) * | 2015-09-18 | 2017-03-23 | Spangler Scientific Llc | Non-invasive determination of risk for sudden cardiac death |
CN105411578A (en) * | 2016-01-02 | 2016-03-23 | 无锡桑尼安科技有限公司 | Automatic patient state detection method |
CN105640521A (en) * | 2016-01-02 | 2016-06-08 | 无锡桑尼安科技有限公司 | Intelligent physiological parameter detection method |
CN106073765A (en) * | 2016-05-27 | 2016-11-09 | 成都信汇聚源科技有限公司 | Die suddenly in a kind of miniature holter devices the implementation method of risk profile |
CN106108889A (en) * | 2016-07-20 | 2016-11-16 | 杨平 | Electrocardiogram classification method based on degree of depth learning algorithm |
CN106214123A (en) * | 2016-07-20 | 2016-12-14 | 杨平 | A kind of electrocardiogram compressive classification method based on degree of depth learning algorithm |
CN106343976A (en) * | 2016-09-14 | 2017-01-25 | 京东方科技集团股份有限公司 | Method and device for establishing blood pressure model as well as method and device for determining blood pressure |
CN107951485A (en) * | 2017-11-27 | 2018-04-24 | 乐普(北京)医疗器械股份有限公司 | Ambulatory ECG analysis method and apparatus based on artificial intelligence self study |
CN108095708A (en) * | 2018-01-19 | 2018-06-01 | 动析智能科技有限公司 | A kind of physiology monitoring and analysis method, system based on mixing sensing |
Cited By (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110037680A (en) * | 2019-03-12 | 2019-07-23 | 平安科技(深圳)有限公司 | The recognition methods of electrocardiogram and device, storage medium, computer equipment |
CN110897630A (en) * | 2019-06-27 | 2020-03-24 | 上海数创医疗科技有限公司 | Ventricular tachycardia and heart rhythm identification method and device based on transfer learning |
WO2021037102A1 (en) * | 2019-08-27 | 2021-03-04 | 中科麦迪人工智能研究院(苏州)有限公司 | Electrocardiogram analysis method and apparatus based on picture and heartbeat information, and device and medium |
CN110693486B (en) * | 2019-09-27 | 2022-06-14 | 武汉中旗生物医疗电子有限公司 | Electrocardiogram abnormity labeling method and device |
CN110693486A (en) * | 2019-09-27 | 2020-01-17 | 武汉中旗生物医疗电子有限公司 | Electrocardiogram abnormity labeling method and device |
CN110720894A (en) * | 2019-11-08 | 2020-01-24 | 广州视源电子科技股份有限公司 | Atrial flutter detection method, device, equipment and storage medium |
CN110693491A (en) * | 2019-11-13 | 2020-01-17 | 歌尔科技有限公司 | Electrocardiosignal processing method and device, wearable device and readable storage medium |
CN111265194A (en) * | 2019-12-11 | 2020-06-12 | 广州视源电子科技股份有限公司 | Ventricular hypertrophy detection method and device, storage medium and processor |
CN111265194B (en) * | 2019-12-11 | 2023-02-14 | 广州视源电子科技股份有限公司 | Ventricular hypertrophy detection method and device, storage medium and processor |
CN111134662A (en) * | 2020-02-17 | 2020-05-12 | 武汉大学 | Electrocardio abnormal signal identification method and device based on transfer learning and confidence degree selection |
CN111134662B (en) * | 2020-02-17 | 2021-04-16 | 武汉大学 | Electrocardio abnormal signal identification method and device based on transfer learning and confidence degree selection |
CN111466911A (en) * | 2020-04-13 | 2020-07-31 | 徐州工业职业技术学院 | Electronic signal processing system and method |
CN111513702A (en) * | 2020-04-30 | 2020-08-11 | 何中煜 | Method for detecting electrocardiographic data and storage medium |
CN114305446A (en) * | 2021-12-25 | 2022-04-12 | 肇庆星网医疗科技有限公司 | Atrial fibrillation detection method and system based on artificial intelligence |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109276242A (en) | The method and apparatus of electrocardiosignal type identification | |
CN106108889B (en) | Electrocardiogram classification method based on deep learning algorithm | |
CN102835954B (en) | Method for generating heart beating waveform template and treating ventricular premature beat detection and device | |
CN108836302B (en) | Intelligent electrocardiogram analysis method and system based on deep neural network | |
Park et al. | Hierarchical support vector machine based heartbeat classification using higher order statistics and hermite basis function | |
CN107837082A (en) | Electrocardiogram automatic analysis method and device based on artificial intelligence self study | |
CN109117730A (en) | Electrocardiogram auricular fibrillation real-time judge method, apparatus, system and storage medium | |
CN104473629B (en) | Automatic electrocardioelectrode placement error detection method based on kernel function classification algorithm | |
CN108056773A (en) | Based on the algorithms of QRS complexes detection in electrocardiogram signal for improving variation mode decomposition | |
CN106725420A (en) | VPB recognition methods and VPB identifying system | |
Zhang et al. | ECG quality assessment based on a kernel support vector machine and genetic algorithm with a feature matrix | |
WO2021071688A1 (en) | Systems and methods for reduced lead electrocardiogram diagnosis using deep neural networks and rule-based systems | |
CN112426160A (en) | Electrocardiosignal type identification method and device | |
CN103637795B (en) | Automatic diagnosis function detection method for electrocardiogram instrument | |
CN104840186A (en) | Evaluation method of autonomic nervous function of patient suffering from CHF (Congestive Heart-Failure) | |
CN111956208B (en) | ECG signal classification method based on ultra-lightweight convolutional neural network | |
CN106874722A (en) | A kind of personal identification method and its device based on electrocardiosignal | |
CN113057648A (en) | ECG signal classification method based on composite LSTM structure | |
CN111481192A (en) | Electrocardiosignal R wave detection method based on improved U-Net | |
CN104834921A (en) | Electrocardio normality/abnormality big-data processing method and device | |
CN110897634A (en) | Electrocardiosignal generation method based on generation countermeasure network | |
CN110236529A (en) | A kind of multi-lead arrhythmia cordis intelligent diagnosing method based on MODWT and LSTM | |
CN112022141A (en) | Electrocardiosignal category detection method and device and storage medium | |
CN115486855A (en) | Electrocardiogram heart beat classification method based on QRS wave group indefinite cycle superposition | |
CN115392319A (en) | Electrocardio abnormality classification method fusing heart dynamics model and antagonistic generation network |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190129 |
|
RJ01 | Rejection of invention patent application after publication |