CN108416277A - A kind of cardioelectric monitor method, apparatus, terminal and computer readable storage medium - Google Patents
A kind of cardioelectric monitor method, apparatus, terminal and computer readable storage medium Download PDFInfo
- Publication number
- CN108416277A CN108416277A CN201810144004.5A CN201810144004A CN108416277A CN 108416277 A CN108416277 A CN 108416277A CN 201810144004 A CN201810144004 A CN 201810144004A CN 108416277 A CN108416277 A CN 108416277A
- Authority
- CN
- China
- Prior art keywords
- qrs complex
- cardioelectric monitor
- waves
- electrocardiosignal
- neural network
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
- G06F2218/10—Feature extraction by analysing the shape of a waveform, e.g. extracting parameters relating to peaks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H40/00—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/02—Preprocessing
- G06F2218/04—Denoising
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- General Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Epidemiology (AREA)
- Computational Linguistics (AREA)
- Business, Economics & Management (AREA)
- Signal Processing (AREA)
- Medical Informatics (AREA)
- Primary Health Care (AREA)
- Public Health (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biophysics (AREA)
- General Business, Economics & Management (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention discloses a kind of cardioelectric monitor method, apparatus, terminal and computer readable storage medium, the method includes：Receive the electrocardiosignal acquired in real time that cardioelectric monitor equipment is sent；To the electrocardiosignal be filtered and denoising after, detect the R waves of the electrocardiosignal, and QRS complex is extracted according to the R waves；The QRS complex is inputted in trained neural network model in advance and is identified；When identifying that the QRS complex is not normal, then the QRS complex not normal to this is classified, and obtains the corresponding arrhythmia cordis type of the not normal QRS complex；Cardioelectric monitor report is generated according to the arrhythmia cordis type, the cardioelectric monitor is exported and reports and trigger alarm, the ecg wave form even to note abnormalities then triggers alarm.The present invention can reduce monitoring operation difficulty, improves monitoring accuracy and realize round-the-clock monitoring electrocardio-activity.
Description
Technical field
The present invention relates to a kind of monitoring device technical field more particularly to cardioelectric monitor method, apparatus, terminal and computers
Readable storage medium storing program for executing.
Background technology
Traditional cardioelectric monitor method need by conducting wire the equipment of the collected signal transmission of electrode to collecting cassette into
Row identifying processing.If it is by multiple electrodes, multiple leads can be corresponding with a plurality of come the cardioelectric monitor realized, then multiple electrodes
Conducting wire, and it is more demanding to the stickup orientation of each electrode, it is inconvenient to use.
Although wireless mode connection electrocardiogram monitoring terminal can be used to realize that long-range electrocardio is supervised in current cardioelectric monitor method
It surveys.But current remote ecg monitoring technology and immature, monitoring accuracy is relatively low, and is related to the complicated for operation of monitoring device,
Popular use is not easy to carry out round-the-clock monitoring electrocardio-activity.
Invention content
Technical problem to be solved of the embodiment of the present invention is, provides a kind of cardioelectric monitor method, apparatus, terminal and meter
Calculation machine readable storage medium storing program for executing can reduce monitoring operation difficulty, improves monitoring accuracy and realize round-the-clock monitoring electrocardio-activity.
In order to solve the above-mentioned technical problem, an embodiment of the present invention provides a kind of cardioelectric monitor methods, include at least as follows
Step：
Receive the electrocardiosignal acquired in real time that cardioelectric monitor equipment is sent；
To the electrocardiosignal be filtered and denoising after, detect the R waves of the electrocardiosignal, and according to the R
Wave extracts QRS complex；
It will be identified in QRS complex input neural network model trained in advance；
When identifying that the QRS complex is not normal, then the QRS complex not normal to this is classified, and obtains the not normal QRS
The corresponding arrhythmia cordis type of wave group；
Cardioelectric monitor report is generated according to the arrhythmia cordis type, the cardioelectric monitor is exported and reports and trigger alarm.
Further, the training step of the neural network model is：
Receive the electrocardiosignal acquired in real time that cardioelectric monitor equipment is sent；
To the electrocardiosignal be filtered and denoising after, detect the R waves of the electrocardiosignal, and according to the R
Wave extracts QRS complex；
The QRS complex is input in neural network the training that exercises supervision, obtains Neural Network Training Parameter；
It is built according to the Neural Network Training Parameter and preserves neural network model.
Further, cardioelectric monitor report includes arrhythmia cordis type and sends out alarm for triggering monitoring terminal moving
The instruction of work；The QRS complex be centered on R waves, respectively forwardly take 40 data and take backward the formation of 169 data with
Sample data of 210 data as feature；The electrocardiosignal includes at least one QRS complex and P waves, U waves and T waves.
Further, the neural network model includes three layers of LSTM networks；Each layer LSTM networks include containing
The visible layer of 70 neurons, the hidden layer containing 200 neurons and the output layer containing 14 neurons；The output layer
14 neurons represent 14 kinds of arrhythmia cordis types.
Further, it will be identified in QRS complex input neural network model trained in advance, specifically
For：
By the QRS complex input build in advance neural network model in be trained, and training complete
Preservation model is used for the identification and classification of the following electrocardiosignal afterwards.
Further, the electrocardio equipment includes at least one electrode and electrocardio processing chip；The electrocardio handles core
Piece with monitoring terminal for wirelessly being communicated.
The embodiment of the present invention additionally provides a kind of electrocardiograph monitoring device, including：
Electrocardiosignal receiving unit, the electrocardiosignal acquired in real time for receiving the transmission of cardioelectric monitor equipment；
QRS complex extraction unit, for the electrocardiosignal be filtered and denoising after, detect electrocardio letter
Number R waves, and according to the R waves extract QRS complex；
QRS complex recognition unit, for will be carried out in QRS complex input neural network model trained in advance
Identification；
Judging unit, for when identifying that the QRS complex is not normal, then the QRS complex not normal to this to be classified, and is obtained
The corresponding arrhythmia cordis type of the QRS complex not normal to this；
Alarm unit exports the cardioelectric monitor report for generating cardioelectric monitor report according to the arrhythmia cordis type
It accuses and triggers alarm.
Further, cardioelectric monitor report includes arrhythmia cordis type and sends out alarm for triggering monitoring terminal moving
The instruction of work；The QRS complex be centered on R waves, respectively forwardly take 40 data and take backward the formation of 169 data with
Sample data of 210 data as feature；The electrocardiosignal includes at least one QRS complex and P waves, U waves and T waves；
The neural network model includes three layers of LSTM networks；Each layer LSTM networks include containing 70 neurons
Visible layer, the hidden layer containing 200 neurons and the output layer containing 14 neurons；14 neurons of the output layer
Represent 14 kinds of arrhythmia cordis types.
The embodiment of the present invention additionally provides a kind of electrocardiogram monitoring terminal, including processor, memory and is stored in described
In memory and it is configured as the computer program that the processor executes, when the processor executes the computer program
Realize cardioelectric monitor method as described in claim 1.
The embodiment of the present invention additionally provides a kind of computer readable storage medium, and the computer readable storage medium includes
The computer program of storage, wherein control when the computer program is run and set where the computer readable storage medium
It is standby to execute cardioelectric monitor method as described in claim 1.
Implement the embodiment of the present invention, has the advantages that：
A kind of cardioelectric monitor method, apparatus, terminal and computer readable storage medium provided in an embodiment of the present invention, it is described
Method includes：Receive the electrocardiosignal acquired in real time that cardioelectric monitor equipment is sent；The electrocardiosignal is filtered and is gone
It makes an uproar after processing, detects the R waves of the electrocardiosignal, and QRS complex is extracted according to the R waves；QRS complex input is advance
It is identified in trained neural network model；When identifying that the QRS complex is not normal, then the QRS wave not normal to this
Group classifies, and obtains the corresponding arrhythmia cordis type of the not normal QRS complex；Electrocardio is generated according to the arrhythmia cordis type
Monitoring report exports the cardioelectric monitor and reports and trigger alarm, and the ecg wave form even to note abnormalities then triggers alarm.This hair
It is bright to reduce monitoring operation difficulty, improve monitoring accuracy and realize round-the-clock monitoring electrocardio-activity.
Description of the drawings
Fig. 1 is a kind of flow diagram for cardioelectric monitor method that first embodiment of the invention provides；
Fig. 2 is the schematic diagram of single QRS complex form in first embodiment of the invention；
Fig. 3 is the schematic diagram of LSTM network structures in first embodiment of the invention；
Fig. 4 is the flow diagram that neural network model is built in first embodiment of the invention；
Fig. 5 is a kind of structural schematic diagram for electrocardiograph monitoring device that second embodiment of the invention provides.
Specific implementation mode
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation describes, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
It should be noted that the monitoring terminal in the embodiment of the present invention can be mobile terminal and immobile terminal, it is non-
Mobile terminal includes desktop computer, and mobile terminal includes smart mobile phone (Smart Phone, such as Android phone, IOS mobile phones
Deng), intelligent glasses, smartwatch, Intelligent bracelet, tablet computer, laptop, personal digital assistant etc. can carry out it is wireless
The mobile internet device of communication.
It should be noted that electrocardiogram is made of a series of wave group, each wave group represents each week aroused in interest
Phase.One wave group includes that P waves, QRS complex, T involve U waves.Electrocardiogram is seen first it is to be understood that meaning representated by each wave.
P waves：Sinoatrial node is risen in the excitement of heart, and then conduction reaches atrium.P waves are every produced by Atrial depolarization
First wave in one wave group, it reflects the process of depolarization of left and right atrium.First half represents right room, and latter half represents a left side
Room.
QRS complex：Typical QRS complex includes three closely coupled waves, and first downward wave is known as Q waves, after Q waves
The upright wave of afterwards one high point is known as R waves, and the wave after R waves under is known as S waves.Because its is closely coupled, and reflect ventricle electricity
Ignition process, therefore it is referred to as QRS complex.This wave group reflects left and right biventricular process of depolarization.
T waves：T waves are located at after S-T segment, are that one relatively low and while accounting for longer wave, it is caused by ventricular bipolar.
U waves：U waves are located at after T waves, relatively low small, and genesis mechanism is not completely clear.It is commonly considered as myocardial activation
" current potential after swashing ".
First embodiment of the invention：
Referring to Fig. 1, Fig. 1 is a kind of flow diagram for cardioelectric monitor method that first embodiment of the invention provides.Institute
Cardioelectric monitor method is stated, following steps are included at least：
S101, the electrocardiosignal acquired in real time that cardioelectric monitor equipment is sent is received.
In the present embodiment, since the cardioelectric monitor equipment can be communicated wirelessly with monitoring terminal, nothing
Conducting wire is needed to connect, so the cardioelectric monitor equipment can be arranged to coin-size according to actual demand and be packaged with electrode, sensing
Device and electrocardio processing chip.The cardioelectric monitor equipment can utilize special sticker or Surgidressing to be pasted onto front, and pass through nothing
Line mode connects monitoring terminal, with the real-time electrocardiosignal for sending acquisition to the monitoring terminal.
S102, to the electrocardiosignal be filtered and denoising after, detect the R waves of the electrocardiosignal, and according to
The R waves extract QRS complex.
In the present embodiment, the monitoring terminal can be mobile phone terminal, and mobile phone terminal is loaded with cardioelectric monitor APP up and down,
The cardioelectric monitor APP is packaged with the neural network handled ECG's data compression using deep learning intelligent recognition algorithm
Model.After the deep learning intelligent recognition algorithm process, electrocardio report is generated and by the cardioelectric monitor report output
The monitoring terminal, i.e. mobile phone terminal are shown and are reminded.
Specifically, the cardioelectric monitor equipment can be the electrocardio device of three leads, for acquiring electrocardiosignal, then will
Collected electrocardiosignal is sent to monitoring terminal.Electrocardiosignal described in the monitoring terminal-pair is filtered with after denoising, is examined
Survey the R waves in the electrocardiosignal.Centered on R waves, 40 data are taken forward, are taken 169 data backward, are then calculated whole
Phase between average between the two neighboring R waves of waveform, and using the phase between average between the two neighboring R waves as QRS complex
A feature, in this way, each QRS complex can use 210 data as feature.Referring to Fig. 2, Fig. 2 is the present invention first
The schematic diagram of single QRS complex form in embodiment.
S103, it will be identified in QRS complex input neural network model trained in advance.
In the present embodiment, referring to Fig. 3, Fig. 3 is the schematic diagram of LSTM network structures in first embodiment of the invention.Institute
State neural network model include three layers of LSTM networks, each layer LSTM networks include the visible layer containing 70 neurons,
Hidden layer containing 200 neurons and the output layer containing 14 neurons.14 neurons of the output layer represent 14 kinds
Arrhythmia cordis type.
It should be noted that LSTM (Long Short-Term Memory) network is shot and long term memory network, it is a kind of
Time recurrent neural network is suitable for being spaced and postpone relatively long critical event in processing and predicted time sequence.LSTM
Three doors of Web vector graphic come control signal input, output.Input gate be used for control currently input number, forget door be used for control
How much the information at moment enters current input before system, and out gate is used for controlling the output at current time.
Referring to Fig. 4, Fig. 4 is the flow diagram that neural network model is built in first embodiment of the invention.The god
Construction step through network model is：
S201, the electrocardiosignal acquired in real time that cardioelectric monitor equipment is sent is received.
S202, to the electrocardiosignal be filtered and denoising after, detect the R waves of the electrocardiosignal, and according to
The R waves extract QRS complex.
S203, the QRS complex is input in neural network the training that exercises supervision, obtains Neural Network Training Parameter.
In the present embodiment, all QRS complexes of whole waveform are extracted, put on label for each QRS complex, often
A label corresponds to the type of each QRS complex.The label or type represent cardioelectric monitor to electrocardiosignal belong to just
Normal atrial fibrillation or other types.
It is understood that N number of QRS complex and corresponding N can be extracted from a waveform of the electrocardiosignal
A label, and the QRS complex is inputted into neural network the training that exercises supervision, Neural Network Training Parameter is obtained, and preserve
Neural Network Training Parameter after training completion, i.e. neural network model parameter.
It should be noted that the QRS complex is centered on R waves, respectively forwardly takes 40 data and take 169 backward
The sample data using 210 data as feature that data are formed；The electrocardiosignal includes at least one QRS complex and P waves,
U waves and T waves.
S204, it is built according to the Neural Network Training Parameter and preserves neural network model.
New electrocardiosignal can be identified and be classified using the neural network model and judge and export electrocardio prison
Observe and predict announcement.
S104, when identifying that the QRS complex is not normal, then the QRS complex not normal to this is classified, and it is not normal to obtain this
The corresponding arrhythmia cordis type of QRS complex.
In the present embodiment, since the LSTM networks in the neural network model obtain 14 neurons of the output layer
Represent 14 kinds of arrhythmia cordis types.It is when the neural network model identifies that the QRS complex is not normal, then not normal to this
QRS complex is classified, and the corresponding arrhythmia cordis type of the not normal QRS complex is obtained, such as：The electrocardio letter that cardioelectric monitor arrives
Number belong to normal atrial fibrillation or other types.
S105, cardioelectric monitor report is generated according to the arrhythmia cordis type, export the cardioelectric monitor and report and trigger
Alarm.
It is understood that the cardioelectric monitor report includes arrhythmia cordis type and sends out police for triggering monitoring terminal
The instruction of report action.I.e. if identifying that the QRS complex is not normal, and judge the corresponding arrhythmia cordis class of the electrocardiosignal
Type generates cardioelectric monitor report, then monitoring terminal can be while showing the cardioelectric monitor report, by triggering jingle bell or shake
Flowing mode reminds user.
In the present embodiment, the cardioelectric monitor report of output will be directly displayed on the screen of the monitoring terminal,
As mobile phone terminal screen on.It is understood that the monitoring terminal, which can also report the cardioelectric monitor, is uploaded to clothes
Business device preserves.
Further, it will be identified in QRS complex input neural network model trained in advance, specifically
For：
It will be trained in QRS complex input neural network model trained in advance, and after the completion of training
The corresponding electrocardiosignal of the QRS complex is identified using deep learning algorithm.
In the present embodiment, it for example, in the CNN neural networks that image processing field is good at very much, uses " local receptor field "
The concept of " power is planted shared ", can greatly reduce the quantity of network parameter.Key is that this structure meets vision very much really
Operation principle of the generic task on human brain.
Further, the electrocardio equipment includes at least one electrode and electrocardio processing chip；The electrocardio handles core
Piece with monitoring terminal for wirelessly being communicated.
A kind of cardioelectric monitor method provided in this embodiment, including：Receive that cardioelectric monitor equipment sends acquires in real time
Electrocardiosignal；To the electrocardiosignal be filtered and denoising after, detect the R waves of the electrocardiosignal, and according to the R
Wave extracts QRS complex；It will be identified in QRS complex input neural network model trained in advance；When identifying
When the QRS complex is not normal, then the QRS complex not normal to this is classified, and is obtained the corresponding heart rate of the not normal QRS complex and is lost
Normal type；Cardioelectric monitor report is generated according to the arrhythmia cordis type, the cardioelectric monitor is exported and reports and trigger alarm, i.e.,
Alarm is triggered if the ecg wave form to note abnormalities.The present invention can reduce monitoring operation difficulty, improve monitoring accuracy and realization
Round-the-clock monitoring electrocardio-activity.
Second embodiment of the invention：
Referring to Fig. 5, Fig. 5 is a kind of structural schematic diagram for electrocardiograph monitoring device that second embodiment of the invention provides.
The present embodiment additionally provides a kind of electrocardiograph monitoring device, including：
Electrocardiosignal receiving unit 301, the electrocardiosignal acquired in real time for receiving the transmission of cardioelectric monitor equipment.
In the present embodiment, since the cardioelectric monitor equipment can be communicated wirelessly with monitoring terminal, nothing
Conducting wire is needed to connect, so the cardioelectric monitor equipment can be arranged to coin-size according to actual demand and be packaged with electrode, sensing
Device and electrocardio processing chip.The cardioelectric monitor equipment can utilize special sticker or Surgidressing to be pasted onto front, and pass through nothing
Line mode connects monitoring terminal, with the real-time electrocardiosignal for sending acquisition to the monitoring terminal.
QRS complex extraction unit 302, for the electrocardiosignal be filtered and denoising after, detect the heart
The R waves of electric signal, and QRS complex is extracted according to the R waves.
In the present embodiment, the monitoring terminal can be mobile phone terminal, and mobile phone terminal is loaded with cardioelectric monitor APP up and down,
The cardioelectric monitor APP is packaged with the neural network handled ECG's data compression using deep learning intelligent recognition algorithm
Model.After the deep learning intelligent recognition algorithm process, electrocardio report is generated and by the cardioelectric monitor report output
The monitoring terminal, i.e. mobile phone terminal are shown and are reminded.
Specifically, the cardioelectric monitor equipment can be the electrocardio device of three leads, for acquiring electrocardiosignal, then will
Collected electrocardiosignal is sent to monitoring terminal.Electrocardiosignal described in the monitoring terminal-pair is filtered with after denoising, is examined
Survey the R waves in the electrocardiosignal.Centered on R waves, 40 data are taken forward, are taken 169 data backward, are then calculated whole
Phase between average between the two neighboring R waves of waveform, and using the phase between average between the two neighboring R waves as QRS complex
A feature, in this way, each QRS complex can use 210 data as feature.Referring to Fig. 2, Fig. 2 is the present invention first
The schematic diagram of single QRS complex form in embodiment.
QRS complex recognition unit 303, for inputting the QRS complex in neural network model trained in advance
It is identified.
In the present embodiment, referring to Fig. 3, Fig. 3 is the schematic diagram of LSTM network structures in first embodiment of the invention.Institute
State neural network model include three layers of LSTM networks, each layer LSTM networks include the visible layer containing 70 neurons,
Hidden layer containing 200 neurons and the output layer containing 14 neurons.14 neurons of the output layer represent 14 kinds
Arrhythmia cordis type.
It should be noted that LSTM (Long Short-Term Memory) network is shot and long term memory network, it is a kind of
Time recurrent neural network is suitable for being spaced and postpone relatively long critical event in processing and predicted time sequence.LSTM
Three doors of Web vector graphic come control signal input, output.Input gate be used for control currently input number, forget door be used for control
How much the information at moment enters current input before system, and out gate is used for controlling the output at current time.
Referring to Fig. 4, Fig. 4 is the flow diagram that neural network model is built in first embodiment of the invention.The god
Construction step through network model is：
S201, the electrocardiosignal acquired in real time that cardioelectric monitor equipment is sent is received.
S202, to the electrocardiosignal be filtered and denoising after, detect the R waves of the electrocardiosignal, and according to
The R waves extract QRS complex.
S203, the QRS complex is input in neural network the training that exercises supervision, obtains Neural Network Training Parameter.
In the present embodiment, all QRS complexes of whole waveform are extracted, put on label for each QRS complex, often
A label corresponds to the type of each QRS complex.The label or type represent cardioelectric monitor to electrocardiosignal belong to just
Normal atrial fibrillation or other types.
It is understood that N number of QRS complex and corresponding N can be extracted from a waveform of the electrocardiosignal
A label, and the QRS complex is inputted into neural network the training that exercises supervision, Neural Network Training Parameter is obtained, and preserve
Neural Network Training Parameter after training completion, i.e. neural network model parameter.
It should be noted that the QRS complex is centered on R waves, respectively forwardly takes 40 data and take 169 backward
The sample data using 210 data as feature that data are formed；The electrocardiosignal includes at least one QRS complex and P waves,
U waves and T waves.
S204, it is built according to the Neural Network Training Parameter and preserves neural network model.
New electrocardiosignal can be identified and be classified using the neural network model and judge and export electrocardio prison
Observe and predict announcement.
Judging unit 304, for when identifying that the QRS complex is not normal, then the QRS complex not normal to this to divide
Class obtains the corresponding arrhythmia cordis type of the not normal QRS complex.
In the present embodiment, since the LSTM networks in the neural network model obtain 14 neurons of the output layer
Represent 14 kinds of arrhythmia cordis types.It is when the neural network model identifies that the QRS complex is not normal, then not normal to this
QRS complex is classified, and the corresponding arrhythmia cordis type of the not normal QRS complex is obtained, such as：The electrocardio letter that cardioelectric monitor arrives
Number belong to normal atrial fibrillation or other types.
It should be noted that the neural network model includes three layers of LSTM networks；Each layer LSTM networks include
Visible layer containing 70 neurons, the hidden layer containing 200 neurons and the output layer containing 14 neurons；The output
14 neurons of layer represent 14 kinds of arrhythmia cordis types.
Alarm unit 305 exports the cardioelectric monitor for generating cardioelectric monitor report according to the arrhythmia cordis type
It reports and triggers alarm.
It is understood that the cardioelectric monitor report includes arrhythmia cordis type and sends out police for triggering monitoring terminal
The instruction of report action.I.e. if identifying that the QRS complex is not normal, and judge the corresponding arrhythmia cordis class of the electrocardiosignal
Type generates cardioelectric monitor report, then monitoring terminal can be while showing the cardioelectric monitor report, by triggering jingle bell or shake
Flowing mode reminds user.
In the present embodiment, the cardioelectric monitor report of output will be directly displayed on the screen of the monitoring terminal,
As mobile phone terminal screen on.It is understood that the monitoring terminal, which can also report the cardioelectric monitor, is uploaded to clothes
Business device preserves.
Further, it will be identified in QRS complex input neural network model trained in advance, specifically
For：
It will be trained in QRS complex input neural network model trained in advance, and after the completion of training
The corresponding electrocardiosignal of the QRS complex is identified using deep learning algorithm.
In the present embodiment, it for example, in the CNN neural networks that image processing field is good at very much, uses " local receptor field "
The concept of " power is planted shared ", can greatly reduce the quantity of network parameter.Key is that this structure meets vision very much really
Operation principle of the generic task on human brain.
Further, the electrocardio equipment includes at least one electrode and electrocardio processing chip；The electrocardio handles core
Piece with monitoring terminal for wirelessly being communicated.
A kind of electrocardiograph monitoring device provided in this embodiment, the heart acquired in real time sent by receiving cardioelectric monitor equipment
Electric signal；To the electrocardiosignal be filtered and denoising after, detect the R waves of the electrocardiosignal, and according to the R waves
Extract QRS complex；It will be identified in QRS complex input neural network model trained in advance；When identifying
State QRS complex it is not normal when, then the QRS complex not normal to this is classified, and obtains the corresponding arrhythmia cordis of not normal QRS complex
Type；Cardioelectric monitor report is generated according to the arrhythmia cordis type, the cardioelectric monitor is exported and reports and trigger alarm, even
The ecg wave form to note abnormalities then triggers alarm.The present invention can reduce monitoring operation difficulty, improve monitoring accuracy and realize complete
Weather monitors electrocardio-activity.
The embodiment of the present invention additionally provides a kind of electrocardiogram monitoring terminal, including processor, memory and is stored in described
In memory and it is configured as the computer program that the processor executes, when the processor executes the computer program
Realize such as above-mentioned cardioelectric monitor method.
The embodiment of the present invention additionally provides a kind of computer readable storage medium, and the computer readable storage medium includes
The computer program of storage, wherein control when the computer program is run and set where the computer readable storage medium
It is standby to execute such as above-mentioned cardioelectric monitor method.
The above is the preferred embodiment of the present invention, it is noted that for those skilled in the art
For, without departing from the principle of the present invention, several improvement and deformations can also be made, these improvement and deformations are also considered as
Protection scope of the present invention.
One of ordinary skill in the art will appreciate that realizing all or part of flow in above-described embodiment method, being can be with
Relevant hardware is instructed to complete by computer program, the program can be stored in a computer read/write memory medium
In, the program is when being executed, it may include such as the flow of the embodiment of above-mentioned each method.Wherein, the storage medium can be magnetic
Dish, CD, read-only memory (Read-Only Memory, ROM) or random access memory (Random Access
Memory, RAM) etc..
Claims (10)
1. a kind of cardioelectric monitor method, which is characterized in that include the following steps：
Receive the electrocardiosignal acquired in real time that cardioelectric monitor equipment is sent；
To the electrocardiosignal be filtered and denoising after, detect the R waves of the electrocardiosignal, and carry according to the R waves
Take QRS complex；
It will be identified in QRS complex input neural network model trained in advance；
When identifying that the QRS complex is not normal, then the QRS complex not normal to this is classified, and obtains the not normal QRS complex
Corresponding arrhythmia cordis type；
Cardioelectric monitor report is generated according to the arrhythmia cordis type, the cardioelectric monitor is exported and reports and trigger alarm.
2. cardioelectric monitor method according to claim 1, which is characterized in that the training step of the neural network model
For：
Receive the electrocardiosignal acquired in real time that cardioelectric monitor equipment is sent；
To the electrocardiosignal be filtered and denoising after, detect the R waves of the electrocardiosignal, and carry according to the R waves
Take QRS complex；
The QRS complex is input in neural network the training that exercises supervision, obtains Neural Network Training Parameter；
It is built according to the Neural Network Training Parameter and preserves neural network model.
3. cardioelectric monitor method according to claim 1 or 2, which is characterized in that the cardioelectric monitor report includes heart rate
Not normal type and the instruction that alarm action is sent out for triggering monitoring terminal；The QRS complex be centered on R waves, respectively to
Before take 40 data and take backward 169 data formed the sample data using 210 data as feature；The electrocardiosignal
Including at least one QRS complex and P waves, U waves and T waves.
4. cardioelectric monitor method according to claim 1 or 2, which is characterized in that the neural network model includes three layers
LSTM networks；Each layer LSTM networks include the visible layer containing 70 neurons, the hidden layer containing 200 neurons and
Output layer containing 14 neurons；14 neurons of the output layer represent 14 kinds of arrhythmia cordis types.
5. cardioelectric monitor method according to claim 1, which is characterized in that train QRS complex input in advance
Neural network model in be identified, specially：
By the QRS complex input build neural network model in be trained, training after the completion of preservation model, it
It may be used for classification and the identification mission of the following electrocardiosignal afterwards.
6. cardioelectric monitor method according to claim 1, which is characterized in that the electrocardio equipment includes at least one electrode
And electrocardio processing chip；The electrocardio processing chip with monitoring terminal for wirelessly being communicated.
7. a kind of electrocardiograph monitoring device, which is characterized in that including：
Electrocardiosignal receiving unit, the electrocardiosignal acquired in real time for receiving the transmission of cardioelectric monitor equipment；
QRS complex extraction unit, for the electrocardiosignal be filtered and denoising after, detect the electrocardiosignal
R waves, and QRS complex is extracted according to the R waves；
QRS complex recognition unit, for will know in QRS complex input neural network model trained in advance
Not；
Judging unit, for when identifying that the QRS complex is not normal, then the QRS complex not normal to this to be classified, and is somebody's turn to do
The corresponding arrhythmia cordis type of not normal QRS complex；
Alarm unit exports the cardioelectric monitor report simultaneously for generating cardioelectric monitor report according to the arrhythmia cordis type
Trigger alarm.
8. electrocardiograph monitoring device according to claim 7, which is characterized in that the cardioelectric monitor report includes arrhythmia cordis
Type and the instruction that alarm action is sent out for triggering monitoring terminal；The QRS complex is respectively forwardly to be taken centered on R waves
40 data and the sample data using 210 data as feature for taking 169 data to be formed backward；The electrocardiosignal includes
At least one QRS complex and P waves, U waves and T waves；
The neural network model includes three layers of LSTM networks；Each layer LSTM networks include containing 70 neurons can
See layer, the hidden layer containing 200 neurons and the output layer containing 14 neurons；14 neurons of the output layer represent
14 kinds of arrhythmia cordis types.
9. a kind of electrocardiogram monitoring terminal, which is characterized in that including processor, memory and be stored in the memory and by
It is configured to the computer program that the processor executes, the processor is realized when executing the computer program as right is wanted
Seek the cardioelectric monitor method described in 1.
10. a kind of computer readable storage medium, which is characterized in that the computer readable storage medium includes the calculating of storage
Machine program, wherein equipment where controlling the computer readable storage medium when the computer program is run is executed as weighed
Profit requires the cardioelectric monitor method described in 1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810144004.5A CN108416277A (en) | 2018-02-11 | 2018-02-11 | A kind of cardioelectric monitor method, apparatus, terminal and computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810144004.5A CN108416277A (en) | 2018-02-11 | 2018-02-11 | A kind of cardioelectric monitor method, apparatus, terminal and computer readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108416277A true CN108416277A (en) | 2018-08-17 |
Family
ID=63128615
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810144004.5A Pending CN108416277A (en) | 2018-02-11 | 2018-02-11 | A kind of cardioelectric monitor method, apparatus, terminal and computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108416277A (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109044340A (en) * | 2018-08-29 | 2018-12-21 | 北京航空航天大学 | Electrocardiogram (ECG) data classification method, device, electronic equipment and storage medium |
| CN109363659A (en) * | 2018-09-30 | 2019-02-22 | 平安科技（深圳）有限公司 | Heart rate monitoring method, apparatus and storage medium based on deep learning |
| CN109431491A (en) * | 2018-09-28 | 2019-03-08 | 上海优加利健康管理有限公司 | A kind of automatic report-generating method and system for cardioelectric monitor |
| CN109871808A (en) * | 2019-02-21 | 2019-06-11 | 天津惊帆科技有限公司 | Atrial fibrillation model training and detecting method and device |
| CN110090016A (en) * | 2019-04-28 | 2019-08-06 | 心医国际数字医疗系统(大连)有限公司 | The method and system of positioning R wave position, the R wave automatic testing method using LSTM neural network |
| CN110141216A (en) * | 2019-05-29 | 2019-08-20 | 清华大学深圳研究生院 | A kind of recognition methods, training method and the system of electrocardiosignal QRS characteristic wave |
| CN110881966A (en) * | 2018-09-10 | 2020-03-17 | 深圳市游弋科技有限公司 | Algorithm for processing electrocardiogram data by using LSTM network |
| WO2020061870A1 (en) * | 2018-09-27 | 2020-04-02 | 深圳大学 | Lstm end-to-end single-lead electrocardiogram classification method |
| CN111184508A (en) * | 2020-01-19 | 2020-05-22 | 武汉大学 | Electrocardiosignal detection device and analysis method based on joint neural network |
| CN112750513A (en) * | 2020-12-31 | 2021-05-04 | 复旦大学附属华山医院 | Parathyroid resection patient management system and method |
| CN113057651A (en) * | 2019-12-30 | 2021-07-02 | 石家庄以岭药业股份有限公司 | Data sending method and device, data acquisition device and storage medium |
| CN115797560A (en) * | 2022-11-28 | 2023-03-14 | 广州市碳码科技有限责任公司 | Head model construction method and system based on near infrared spectrum imaging |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101785670A (en) * | 2009-01-22 | 2010-07-28 | 陈跃军 | Intelligent blurry electrocardiogram on-line analyzer system |
| CN104207769A (en) * | 2014-08-27 | 2014-12-17 | 电子科技大学 | Electrocardiosignal detection system |
| CN104382582A (en) * | 2014-11-10 | 2015-03-04 | 哈尔滨医科大学 | Device for classifying dynamic electrocardio data |
| CN106066933A (en) * | 2016-05-27 | 2016-11-02 | 成都信汇聚源科技有限公司 | The method obtaining atrial fibrillation identification artificial neural network weighted value matrix |
| CN106725426A (en) * | 2016-12-14 | 2017-05-31 | 深圳先进技术研究院 | A kind of method and system of electrocardiosignal classification |
-
2018
- 2018-02-11 CN CN201810144004.5A patent/CN108416277A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101785670A (en) * | 2009-01-22 | 2010-07-28 | 陈跃军 | Intelligent blurry electrocardiogram on-line analyzer system |
| CN104207769A (en) * | 2014-08-27 | 2014-12-17 | 电子科技大学 | Electrocardiosignal detection system |
| CN104382582A (en) * | 2014-11-10 | 2015-03-04 | 哈尔滨医科大学 | Device for classifying dynamic electrocardio data |
| CN106066933A (en) * | 2016-05-27 | 2016-11-02 | 成都信汇聚源科技有限公司 | The method obtaining atrial fibrillation identification artificial neural network weighted value matrix |
| CN106725426A (en) * | 2016-12-14 | 2017-05-31 | 深圳先进技术研究院 | A kind of method and system of electrocardiosignal classification |
Non-Patent Citations (4)
| Title |
|---|
| VYKINTAS MAKNICKAS等: "Atrial Fibrillation Classification Using QRS Complex Features and LSTM", 《COMPUTING IN CARDIOLOGY》 * |
| 周珂等: "基于小波神经网络方法的心电图分类研究", 《微电子学与计算机》 * |
| 王继成等: "基于神经网络的一种心电图分类法", 《中国生物医学工程学报》 * |
| 苏志同等: "基于SVM的心律失常的研究和分析", 《软件》 * |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109044340A (en) * | 2018-08-29 | 2018-12-21 | 北京航空航天大学 | Electrocardiogram (ECG) data classification method, device, electronic equipment and storage medium |
| CN110881966A (en) * | 2018-09-10 | 2020-03-17 | 深圳市游弋科技有限公司 | Algorithm for processing electrocardiogram data by using LSTM network |
| WO2020061870A1 (en) * | 2018-09-27 | 2020-04-02 | 深圳大学 | Lstm end-to-end single-lead electrocardiogram classification method |
| CN109431491A (en) * | 2018-09-28 | 2019-03-08 | 上海优加利健康管理有限公司 | A kind of automatic report-generating method and system for cardioelectric monitor |
| CN109363659A (en) * | 2018-09-30 | 2019-02-22 | 平安科技（深圳）有限公司 | Heart rate monitoring method, apparatus and storage medium based on deep learning |
| CN109871808A (en) * | 2019-02-21 | 2019-06-11 | 天津惊帆科技有限公司 | Atrial fibrillation model training and detecting method and device |
| CN109871808B (en) * | 2019-02-21 | 2021-03-02 | 天津惊帆科技有限公司 | Atrial fibrillation model training and detecting method and device |
| CN110090016B (en) * | 2019-04-28 | 2021-06-25 | 心医国际数字医疗系统(大连)有限公司 | Method and system for positioning R wave position and R wave automatic detection method using LSTM neural network |
| CN110090016A (en) * | 2019-04-28 | 2019-08-06 | 心医国际数字医疗系统(大连)有限公司 | The method and system of positioning R wave position, the R wave automatic testing method using LSTM neural network |
| CN110141216A (en) * | 2019-05-29 | 2019-08-20 | 清华大学深圳研究生院 | A kind of recognition methods, training method and the system of electrocardiosignal QRS characteristic wave |
| CN113057651A (en) * | 2019-12-30 | 2021-07-02 | 石家庄以岭药业股份有限公司 | Data sending method and device, data acquisition device and storage medium |
| CN111184508A (en) * | 2020-01-19 | 2020-05-22 | 武汉大学 | Electrocardiosignal detection device and analysis method based on joint neural network |
| CN112750513A (en) * | 2020-12-31 | 2021-05-04 | 复旦大学附属华山医院 | Parathyroid resection patient management system and method |
| CN112750513B (en) * | 2020-12-31 | 2024-04-05 | 复旦大学附属华山医院 | Parathyroid resection patient management system and method |
| CN115797560A (en) * | 2022-11-28 | 2023-03-14 | 广州市碳码科技有限责任公司 | Head model construction method and system based on near infrared spectrum imaging |
| CN115797560B (en) * | 2022-11-28 | 2023-07-25 | 广州市碳码科技有限责任公司 | Near infrared spectrum imaging-based head model construction method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108416277A (en) | A kind of cardioelectric monitor method, apparatus, terminal and computer readable storage medium | |
| CA3071699C (en) | Detection of electrocardiographic signal | |
| US20170112401A1 (en) | Automatic method to delineate or categorize an electrocardiogram | |
| CN1909831B (en) | Cardiac monitoring method, cardiac monitoring system and device | |
| US20190223739A1 (en) | An automatic method to delineate or categorize an electrocardiogram | |
| US11457853B2 (en) | Methods and systems for automatically detecting events based on ECG signals determined from compressed sensed measurements | |
| US11331034B2 (en) | Automatic method to delineate or categorize an electrocardiogram | |
| CN111345816B (en) | Multi-lead QRS complex detection method, device, equipment and storage medium | |
| CN107341880A (en) | A kind of home security method and system | |
| US20070112276A1 (en) | Differentiating arrhythmic events having different origins | |
| CN103690156A (en) | Heart rate acquiring method and electrocardiosignal processing method | |
| CN110236520A (en) | ECG type recognition methods and device based on double convolutional neural networks | |
| CN110638430B (en) | Method for building cascade neural network ECG signal arrhythmia classification model | |
| CN108601540A (en) | Auricular fibrillation detecting system and application method | |
| CN104207769A (en) | Electrocardiosignal detection system | |
| CN109431491A (en) | A kind of automatic report-generating method and system for cardioelectric monitor | |
| CN107865655A (en) | Ecg wave form figure display methods and device | |
| US11571161B2 (en) | Systems and methods for electrocardiogram diagnosis using deep neural networks and rule-based systems | |
| EP3641628B1 (en) | System and method for classifying ectopic cardiac beats | |
| CN107844338A (en) | Application program management-control method, device, medium and electronic equipment | |
| CN110693485A (en) | Wearable electrocardiogram information monitoring system and method | |
| US20240099665A1 (en) | Electrocardiogram data processing server, electrocardiogram data processing method of extracting analysis required section while segmenting electrocardiogram signal into signal segments with variable window sizes, and computer program | |
| CN109480828A (en) | A kind of ecg information processing method, device, computer installation and storage medium | |
| CN109328029B (en) | Vital sign data statistical system and monitor | |
| CN202365773U (en) | Remote wireless electrocardio real-time monitoring system |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180817 |