CN108335744B - A kind of emergency cardiovascular care network system and its method for early warning of classifying - Google Patents
A kind of emergency cardiovascular care network system and its method for early warning of classifying Download PDFInfo
- Publication number
- CN108335744B CN108335744B CN201810289623.3A CN201810289623A CN108335744B CN 108335744 B CN108335744 B CN 108335744B CN 201810289623 A CN201810289623 A CN 201810289623A CN 108335744 B CN108335744 B CN 108335744B
- Authority
- CN
- China
- Prior art keywords
- data
- network
- ecg
- matrix
- early warning
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Classifications
-
- 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
- G16H40/60—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 for the operation of medical equipment or devices
- G16H40/67—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 for the operation of medical equipment or devices for remote operation
-
- 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
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
-
- 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
- G16H80/00—ICT specially adapted for facilitating communication between medical practitioners or patients, e.g. for collaborative diagnosis, therapy or health monitoring
-
- 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
- G06F2218/08—Feature extraction
-
- 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
- Health & Medical Sciences (AREA)
- Engineering & Computer Science (AREA)
- Biomedical Technology (AREA)
- Medical Informatics (AREA)
- Public Health (AREA)
- General Health & Medical Sciences (AREA)
- Epidemiology (AREA)
- Primary Health Care (AREA)
- Pathology (AREA)
- Business, Economics & Management (AREA)
- General Business, Economics & Management (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention discloses a kind of emergency cardiovascular care network system and its classification method for early warning, including doctor's client, first-aid centre's server system, 12 lead electrocardiogram acquisition modules, the information of acquisition is sent to Medical flat computer by bluetooth by the 12 lead electrocardiogram acquisition module, and first-aid centre's server system is sent to by 4G wireless network, first-aid centre's server system is connected by network with doctor's client.The present invention can carry out classification early warning by emergency to patient's anomalous ecg symptom, help Medical Technologist to cardiovascular carly fruit drop, so that the medical treatment time of cardiovascular patient greatly shortens.
Description
Technical field
The invention belongs to medical signals process field more particularly to a kind of emergency cardiovascular care network system and its classification early warning
Method.
Background technique
Acute cardiovascular disease (Acute Cardiovascular Diseases, ACVD) is one group and involves heart and surrounding
The acute clinical syndrome of the circulatory system, the existing cardiovascular patient 2.7 hundred million in China, annual 3000000 people die of ACVD, occupy people
The 30% of the total death of mouth；The number for dying of cardiovascular disease to the year two thousand thirty will be increased to 20,000,000 people；ACVD is further promoted to rescue
It is very urgent to control ability.
Currently, under the existing Medical treatment system in China medical resource and technology distribution extreme imbalance；Medical technique level is low
Under, quality of medical care is not able to satisfy patient's primary demand, and the public lacks necessary vigilance and understanding to the clinical manifestation of ACVD, finally
Cause most of patients that cannot quickly and effectively be given treatment in time window and lethal and disable.Quickly and effectively EARLY RECOGNITION,
It is most important to diagnose ACVD.Electrocardiogram can directly reflect the health status of human heart, and to some cardiovascular diseases such as cardiac muscle
The diagnosis of ischemic, atrioventricular hypertrophy, various arrhythmia cordis has biggish accuracy, but there is following problems: clinician exists
Interpreting blueprints and diagnosis during, need a large amount of waveform parameter of precise measurement, at the same also to have clinical diagnosis experience abundant and
Theoretical knowledge, this certainly will influence correct diagnosis of the medical institutions to clinical heart disease, and experience medical personnel not abundant sentences
It reads difficult and erroneous decision may be made.In addition, being read since the medical personnel for needing professional knowledge manually carries out electrocardiogram
It reads, so being unfavorable for patient carries out familial self detection, is also not easy to the trouble in the mountain area such as having inconvenient traffic to some Special sections
Person makes diagnosis in time.
Internet of Things is the practical platform of new information technology to be got up in recent years based on internet development, can be with by the platform
The common monitoring such as the vital sign by patient, checking information and medical history are real-time transmitted to area medical center by 4G network, in
Heart expert can carry out remote medical consultation with specialists and guidance rescue on multiple computer terminals or mobile phone in institute.The system can cover periphery
Regional base and community hospital when network hospital carries out real-time monitoring to ACVD patient using the system, live can remotely refer to
Lead the patient for the treatment that needs to change the place of examination.Patient care information is uninterruptedly transmitted to region doctor by also mountable system on 120 emergency tenders
Treatment center, expert carry out site assessment to patient, and patient, which enters ambulance, can start pectoralgia center, accomplish ACVD treatment
" seamless link ".
Summary of the invention
Based on problem set forth above, the present invention provides a kind of specification procedures, rapid, easily and effectively painstaking effort
Pipe emergency network system and its classification method for early warning.
In order to realize above method, technical solution used by present system are as follows:
A kind of emergency cardiovascular care network system, including clinician user terminal, first-aid centre's server system, 12 lead electrocardios
The information of acquisition is sent to Medical flat computer by bluetooth by figure acquisition module, the 12 lead electrocardiogram acquisition module, and
First-aid centre's server system is sent to by 4G wireless network, and first-aid centre's server system completes the classification of information data
Processing, and be connected by network with clinician user terminal；
First-aid centre's server system includes database server, network server, data processing module, Yi Shengke
Family end, the database server receive 4G wireless network transmission patient correlation acquisition data after, respectively with network service
Device, data processing module realize interconnection, and the network server, data processing module are connected with doctor's client；The net
For executing control instruction operation and executing inquiry monitoring operation, doctor can check the personal information of patient, rescue network server
Control specific location locating for state, first aid information and patient；The database server is used for transmission, accesses and establish patient
People's information, doctor's essential information, medical monitoring data, alarm parameter setting, idagnostic logout and GPS positioning information；The data
Processing module for electrocardiosignal denoising, extract after denoising the temporal signatures of signal and compression reconfiguration, to the signal after reconstruct point
Class and to common symptom according to emergency carry out early warning.
Further, first-aid centre's server system setting is in primary network station hospital, can obtain primary network station hospital,
Two grade network hospital, three-level network hospital, 120 emergency tenders, high-risk Yi Fa group and the high-risk easy GPS for sending out individual present position are fixed
Position information, and primary network station hospital has highest decision-making power；Level-one can be arranged in the 12 lead electrocardiogram acquisition module
Network hospital perhaps two grade network hospital or three-level network hospital；Or 120 emergency platform or high-risk Yi Fa group and
In high-risk easy hair individual.
Further, the clinician user terminal is patient monitor or medical PDA or mobile phone.
Further, the network server includes login interface, function of tonic chord interface, and exploitation environment is Visual
Studio 2010, dynamic web page interactive mode use asp.net technology；Program uses scripting language combination Ajax technology
The asynchronous refresh page is embedded in ActiveX control to connect doctor's client, executes corresponding control instruction.
Further, the database server runs on win7 environment, using the SQL Server of Microsoft Corporation
2000 data base management systems are as developing instrument, and the VC+6.0 of Microsoft is as database front-end, CPU:AMD XP1800
+, memory kingston 3G DDR, hard disk Dell 600G.
A kind of technical solution of method of the invention are as follows: emergency cardiovascular care network system according to claim 1 point
Class method for early warning, comprising the following steps:
Step 1 is first filtered the low-frequency disturbance that is mingled in common electrocardiosignal, electrode slice contact noise, adopts
Low frequency baseline drift interference is filtered out with median filtering, formula is as follows:
Wherein, x (i) is electrocardiosignal sequence, and k is window width, and C is length of window, and y (i) is the signal after denoising, x
For original signal, y is filtered signal,The respectively average value of original signal and filtered signal；
During being filtered out to high-frequency noise, wavelet coefficient is handled using soft-threshold function, formula is as follows:
Wherein, b, bt, T be respectively electrocardiosignal original wavelet coefficients, treated wavelet coefficient and given threshold value,
Sign () is function, and SNR is signal-to-noise ratio, and MSE is root-mean-square error；
Step 2 extracts its temporal signatures to pretreated electrocardiosignal, generates training sample set, instructs to electrocardiogram (ECG) data
Base for practicing sample set data characteristics carries out sparse coding, determines the number P of dictionary base, at random from electrocardiogram (ECG) data training
Training sample is extracted in sample set forms initial dictionary D0∈RN×P, and the optimization of initial dictionary is carried out, random simulation generates some
Electrocardiogram (ECG) data tests signal f ∈ RM×1, compression drop a dimension u=Φ f is carried out to the test signal, signal finally is tested to electrocardiogram (ECG) data
The reconstruct of f；D0It is sample for dictionary, R, M is the sampling number of each training sample, μ is that electrocardiogram (ECG) data is obtained in sampling
Signal, Φ are observing matrix；
Step 3 carries out classification processing to the electrocardiosignal after reconstruct using feature migration algorithm, show that common electrocardio is different
Normal symptom, and early warning is carried out according to emergency to common anomalous ecg symptom using fuzzy algorithmic approach.
Further, the step 2 the following steps are included:
Step 2.1, it includes between normal P wave, QRS complex, T wave, PR interphase, RR that electrocardiogram (ECG) data signal model is established in emulation
Phase, QT interphase, ST sections；It is E=[E that emulation, which generates a large amount of electrocardiogram (ECG) data signal composition training sample set,1,E1,…,E7]∈RM ×W, wherein W is total training sample number, and M is the sampling number of each training sample, and R is sample；
Step 2.2, it determines the number P of dictionary base, is concentrated from electrocardiogram (ECG) data training sample extract 72 training at random
Sample forms initial dictionary D0∈RN×P, the objective function of initial dictionary optimization are as follows:Wherein A0It is training sample set E in dictionary D0On
Rarefaction representation matrix, λ is regularization parameter；
Step 2.3, observing matrix Φ ∈ RN×M(N < < M) is random using independent identically distributed Gaussian random variable composition
Gaussian matrix is realized, carries out compression dimensionality reduction u=Φ f to the test signal, obtains the low-dimensional compression observation signal u of the test signal
∈RN×1, complete the compression sampling of test signal；
Step 2.4, to the reconstruct of electrocardiogram (ECG) data test signal f, using based on l0Signal reconstruction under norm it is non-convex excellent
The orthogonal matching pursuit algorithm changed in derivation algorithm solves equation Φ Da=u, obtains rarefaction representation of the test signal under dictionary
Matrix a ∈ RP×1；Anti- solution reconstruct electrocardiogram (ECG) data tests signal
Further, the step 3 carries out classification processing to the electrocardiosignal after reconstruct using feature migration algorithm, obtains
The detailed process of common anomalous ecg symptom are as follows:
Input: using the electrocardiogram (ECG) data test signal after reconstruct as source domain data set DS, tally set YS, will be between RR
Phase, QRS complex, P wave, T wave, PR interphase, QT interphase, ST sections be used as target domain data set DT, lower-dimensional subspace dimension k, ginseng
Count α, β, γ, λ, benchmark classifier f, maximum number of iterations Tmax；Output: low dimension projective matrix W1；
Step 3.1: pass through the element of distributional difference matrix:
Construct edge distribution difference matrix L0, construct Laplacian Matrix M1, seek cross-cutting constraint factor di, while setting St、SbIt is 0；
Wherein, nS、nTRespectively former field number of samples and target domain number of samples, xi,xjRespectively sample to be tested, n are number；
Step 3.2:t=1；Wherein, t is the number of iterations；
Step 3.3: solving formulaOrObtain projection matrix W1；Wherein, X=DS∪DT,
C is classification number, LCFor condition distributional difference matrix, I is the matrix that element is 1, and K is iterative parameter, StTotal divergence square between class
Battle array, SbClass scatter matrix,For StNon-linear form；
Step 3.4: by DSAnd DTPass through matrix Z=W1 TX is mapped in k n-dimensional subspace n, obtains ZSAnd ZT, φ is diagonal letter
Number；
Step 3.5: in data set { ZS,YSOn training benchmark classifier f, and using the obtained classifier of training to ZTInto
Line flag obtains the tally set of target domain data
Step 3.6: using data set { XS,YSAndAccording to
Building condition distributional difference matrix LC；Simultaneously
According to formulaWithS is solved respectivelybAnd St,
It is when non-linearWithWherein, NCFor the number of samples of C class, u(c)For the mean value of C classification, SwFor Scatter Matrix in class；
Step 3.7:t ← t+1；
Step 3.8: if t >=TmaxOrNo longer change, exports W1；Otherwise, 3.3 are gone to step.
Further, the step 3 carries out early warning according to emergency to common anomalous ecg symptom using fuzzy algorithmic approach
Process are as follows: stop fighting by sorting out electrocardio, the electrocardio of twist mode ventricular tachycardia, ventricular fibrillation, third degree A-V block and myocardial infarction
Exception Type, is classified into that urgent, moderate is urgent, urgent three obscuring components of severe carry out early warning；XT use of fighting is stopped for electrocardio
Triangleshape grade of membership function Trimf training, for twist mode ventricular tachycardia NS, subordinating degree function is all made of Triangleshape grade of membership function
Trimf training simply divides several sections for ventricular fibrillation SC and determines；For third degree A-V block ZZ, subordinating degree function is equal
Using Gaussian subordinating degree function Gaussmf training, for myocardial infarction GS, subordinating degree function uses bilateral Gaussian degree of membership
Function Gauss2mf training.
The present invention has following technical effect:
The present invention improves sampling efficiency and the time of electrocardiogram (ECG) data, reduces the lengthy and jumbled of data；Pass through electrocardiogram (ECG) data
The feature training study of dictionary is obtained to be adapted with electrocardiogram (ECG) data to be tested, can more accurately be captured in electrocardiosignal
P wave, QRS complex, T wave, PR interphase, RR interphase, ST sections of main feature information can be measured in conjunction with compressive sensing theory from compression
Original signal is accurately reconstructed in signal, improves the sparsity of compression and the accuracy of reconstruct.The present invention is to electrocardiosignal
It using median filtering, chooses optimal length of window and is filtered, filtering out low frequency baseline drift interference can guarantee effectively to remove
The baseline drift close with electrocardiosignal is gone, and it is undistorted to be able to maintain electrocardiosignal detail section；It is handled by soft-threshold to height
Frequency noise is filtered out, and this method has good continuity, and is less prone to signal concussion.Feature migration of the present invention is calculated
Method carries out classification processing to the electrocardiosignal after reconstruct, will using the electrocardiogram (ECG) data of reconstruct test signal as source domain data set
Common anomalous ecg type reduces source domain in proper subspace and arrives with sample in target domain as target domain data set
The distance of other side's average point, so that sample is even closer between field after projection, to the common electrocardio obtained after ECG's data compression
Abnormal symptom nicety of grading is higher.The present invention using fuzzy algorithmic approach by sorting out anomalous ecg type, can be divided into it is urgent, in
Degree is urgent, urgent three obscuring components of severe carry out early warning, may be implemented to common anomalous ecg symptom according to emergency into
The different grades of early warning of row, quickly and effectively EARLY RECOGNITION, diagnosis ACVD and the correct diagnosis to clinical heart disease.
Detailed description of the invention
Fig. 1 is emergency cardiovascular care network system block diagram of the invention；
Fig. 2 is the functional block diagram of first-aid centre's server system of the invention；
Fig. 3 is that pretreated electrocardiosignal extracts its temporal signatures, compression reconfiguration flow chart；
Fig. 4 is fuzzy algorithmic approach subordinating degree function figure；
Fig. 5 is that doctor's client software manages main interface.
Specific embodiment
The embodiment of the present invention is described below in detail, examples of the embodiments are shown in the accompanying drawings, wherein from beginning to end
Same or similar label indicates same or similar element or element with the same or similar functions.Below with reference to attached
The embodiment of figure description is exemplary, and for explaining only the invention, and is not considered as limiting the invention.
Before being illustrated to system and method for the invention, introduce first related to the present invention cardiovascular anxious
Rescue the acquisition and Transmission system, the concept of GPS positioning tracking system of the framework, patient's pathological parameter of network.
The framework of emergency cardiovascular care network
Establish the primary network station hospital centered on a certain Medical Group, each hospital of branch and people from county-level city belonging to group
People hospital is two grade network hospital, the three-level network hospital in community, small towns below each two grade network hospital developing deeply；Network doctor
The treatment network of institute, 120 emergency platforms and high-risk Yi Fa group and high-risk easy hair individual foundation is built into jointly based on shifting
60 kilometers of emergency cardiovascular care networks of animal networking.
The network connection parameter of each rank network hospital, each rank network hospital are set in network connection parameter library first
Setting with corresponding Connecting quantity in network connection parameter library, then pass through 4G wireless network or broadband access.
It is to be noted that primary network station hospital has highest decision-making power, in emergency cardiovascular care network, each second level net
The three-level network hospital in community, small towns below network hospital developing deeply, 120 emergency platforms, high-risk Yi Fa group and high-risk easy hair
Body gives treatment to network and the ACVD indication information of respective collected patient is sent to primary network station hospital.It is all to enter treatment net
The hospital of network, may be implemented remote medical consultation with specialists, quickly change the place of examination, data sharing, medical record management, standardization treatment etc., to realize
Medical resource inside and outside group optimizes, the optimization of minimumization of patients ' expenses and prognosis at a specified future date.
The acquisition and Transmission system of patient's pathological parameter
It is on primary network station hospital, two/three-level network hospital, 120 emergency tenders and high-risk in order to implement emergency cardiovascular care
Residence where Yi Fa group and high-risk easy hair individual is equipped with small in size, portable, high reliablity and implantation low in energy consumption
12 lead electrocardiogram acquisition modules.After morbidity, ECG ST section, left ventricular ejection fraction, the blood pressure of patient are acquired in time
Deng vital sign parameter relevant to conditions of patients.
The transmission bandwidth of this system reaches 1M or more, ensure that mass data is not limited by network speed and postponed, and realizes
Real long-range real-time Transmission；In transmission process, the information that no compression transmission mode ensure that reception terminal is shown is undistorted,
And the data of all storages can call playback, printing and measurement analysis at any time.
This system design and operation is in win7 environment, using 2000 data base administration of SQL Server of Microsoft Corporation
System is as developing instrument, and the VC+6.0 of Microsoft is as database front-end, CPU:AMD XP1800+, memory kingston
3G DDR, hard disk Dell 600G.
GPS positioning tracking system
On primary network station hospital, two/three-level network hospital, 120 emergency tenders and high-risk Yi Fa group and high-risk easy hair
Residence where individual is equipped with GPS positioning tracking system, ACVD immediate care center (first-aid centre's clothes of primary network station hospital
Business device system) two/three-level network hospital, 120 emergency tenders and high-risk Yi Fa group and high-risk easy hair individual can be obtained in real time
The location information in place residence provides navigation position information for subsequent early warning rescue.
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 description.
As shown in Figure 1, system of the invention includes clinician user terminal, first-aid centre's server system, 12 lead electrocardios
The information of acquisition is sent to Medical flat computer by bluetooth by figure acquisition module, the 12 lead electrocardiogram acquisition module, and
First-aid centre's server system is sent to by 4G wireless network；First-aid centre's server system passes through network and clinician user
Terminal is connected, and proposes classification early warning scheme according to conditions of patients situation.
First-aid centre's server system
In conjunction with attached drawing 2, including database server, network server, data processing module, doctor's client, the number
According to library server receive 4G wireless network transmission patient correlation acquisition data after, respectively with network server, data processing
Module realizes interconnection, and the network server, data processing module are connected with doctor's client, separately below with regard to modules
Between the course of work describe in detail.
1, data processing module is realized by following algorithm steps:
1) low-frequency disturbance that is mingled in common electrocardiosignal, electrode slice contact noise are filtered first, in
Value filtering filters out low frequency baseline drift interference, and formula is as follows:
Wherein, x (i) is electrocardiosignal sequence, and k is window width, and C is length of window, and y (i) is the signal after denoising, x
Respectively original signal, y are filtered signal,The respectively average value of original signal and filtered signal；
During being filtered out to high-frequency noise, wavelet coefficient is handled using soft-threshold function, formula is as follows:
Wherein, b, bt, T be respectively electrocardiosignal original wavelet coefficients, treated wavelet coefficient and given threshold value,
SNR is signal-to-noise ratio, and MSE is root-mean-square error.
2) its temporal signatures is extracted to pretreated electrocardiosignal, generates training sample set, to electrocardiogram (ECG) data training sample
Base of this collection data characteristics carries out sparse coding, the number P of dictionary base is determined, at random from electrocardiogram (ECG) data training sample
It concentrates and extracts the initial dictionary D of training sample composition0∈RN×P, and the optimization of initial dictionary is carried out, random simulation generates some electrocardio
Data testing signals f ∈ RM×1, compression dimensionality reduction u=Φ f is carried out to the test signal, finally to the weight of electrocardiogram (ECG) data test signal f
Structure；D0It is sample for dictionary, R, the signal that M is the sampling number of each training sample, μ is electrocardiogram (ECG) data to be obtained in sampling,
Φ is observing matrix.
In conjunction with attached drawing 3, step 2.1, model includes normal P wave, QRS complex, T wave, PR interphase, RR interphase, QT interphase, ST
Section；It is E=[E that emulation, which generates a large amount of electrocardiogram (ECG) data signal composition training sample set,1,E1,…,E7]∈RM×W, wherein W is total
Training sample number, M are the sampling number of each training sample, and R is sample；The fundamental frequency of electrocardiogram (ECG) data signal is 40Hz,
Sampling time is 15 periods, sample rate 6400, M=720；
Step 2.2, it determines the number P of dictionary base, is concentrated from electrocardiogram (ECG) data training sample extract 72 training at random
Sample forms initial dictionary D0∈RN×P, the objective function of initial dictionary optimization are as follows:Wherein A0It is training sample set E in dictionary D0On
Rarefaction representation matrix, λ are regularization parameter, λ=2；Initialization the number of iterations initial value is t=1；In compressed sensing, with observation square
Battle array Φ ∈ RN×MTo test electrocardiogram (ECG) data signal f ∈ RM×1It is observed sampling, wherein N < < M, i.e., the higher-dimension electrocardio number tieed up M
It is believed that number dimensionality reduction is into low-dimensional N-dimensional；Compression sampling procedural representation at this time are as follows:U is electrocardio number in formula
According to the signal obtained in sampling, as observation signal；D∈RM×PIt is the sparse transformation basic matrix of dictionary, a ∈ RP×1It is electrocardio
Rarefaction representation matrix of the data f under dictionary, wherein the most elements of a are approximately zero,It is perception matrix；
Observing matrix Φ ∈ RN×MThe random gaussian matrix that (N < < M) is formed using independent identically distributed Gaussian random variable
It realizes, compression dimensionality reduction u=Φ f is carried out to the test signal, obtain the low-dimensional compression observation signal u ∈ R of the test signalN×1, complete
At the compression sampling of test signal；
Reconstruct to electrocardiogram (ECG) data test signal f, using based on l0The non-convex optimization of signal reconstruction under norm, which solves, to be calculated
Orthogonal matching pursuit algorithm in method solves equation Φ Da=u, obtains rarefaction representation matrix a ∈ R of the test signal under dictionaryP ×1；Anti- solution reconstruct electrocardiogram (ECG) data tests signalIt determines that the sub- number of dictionary base is 72 in embodiment, selects electrocardio number at random
Dictionary, while initialized target function initial value J=50, iteration tolerable error J are initialized according to 72 samples in training samples=
0.02, iteration maximum times m=40, reconstructed error is less than 1.87%.
3) feature migration algorithm carries out classification processing to the electrocardiosignal after reconstruct, obtains the tool of common anomalous ecg symptom
Body process are as follows:
Input: using the electrocardiogram (ECG) data test signal after reconstruct as source domain data set DS, tally set YS, will be between RR
Phase, QRS complex, P wave, T wave, PR interphase, QT interphase, ST sections be used as target domain data set DT, lower-dimensional subspace dimension k, ginseng
Count α, β, γ, λ, benchmark classifier f, maximum number of iterations Tmax；Output: low dimension projective matrix W1；
Step 1: pass through the element of distributional difference matrix:Structure
Build edge distribution difference matrix L0, construct Laplacian Matrix M1, seek cross-cutting constraint factor di, while setting St、SbIt is 0；Its
In, nS、nTRespectively former field number of samples and target domain number of samples, xi,xjRespectively sample to be tested, n are number；
Step 2:t=1；Wherein, t is the number of iterations；
Step 3: solving formulaOrObtain projection matrix W1；Wherein, X=DS∪
DT, C is classification number, LCFor condition distributional difference matrix, I is the matrix that element is 1, and K is iterative parameter, StIt is always dissipated between class
Spend matrix, SbClass scatter matrix,For StNon-linear form；
Step 4: by DSAnd DTPass through matrix Z=W1 TX is mapped in k n-dimensional subspace n, obtains ZSAnd ZT, φ is diagonal function；
Step 5: in data set { ZS,YSOn training benchmark classifier f, and using the obtained classifier of training to ZTIt carries out
Label, obtains the tally set of target domain data
Step 6: using data set { XS,YSAndAccording to
Building condition distributional difference matrix LC；Simultaneously
According to formulaWithS is solved respectivelybAnd St(
It is when non-linearWith),
Wherein, NCFor the number of samples of C class, u(c)For the mean value of C classification, SwFor Scatter Matrix in class；
Step 7:t ← t+1；
Step 8: if t >=TmaxOrNo longer change, exports W1；Otherwise, 3 are gone to step.
In embodiment,
RR wave (i): the interval between the current peak R i and next peak i+1R, second (s)；
HRV wave (i): the difference of current RR interphase i and the absolute value of next i+1RR interphase, second (s)；
QRS complex (i): when front center claps the time limit length of QRS complex in i, second (s)；
P wave (i): when front center claps the time limit length of P wave in i, second (s)；
T wave (i): when front center claps the time limit length of T wave in i, second (s)；
PR wave (i): when front center claps the time limit length of PR interphase in i, second (s)；
QT wave (i): when front center claps the time limit length of QT interphase in i, second (s)；
Tmax=30, k value are 7, and parameter alpha, β, γ, λ are respectively 0.5,0.2,10,80；Former field number of samples nS=
1024, target domain number of samples nT=234, n=1600, the number of iterations t=30 in step, classification number C=in step 3
18。
4) as shown in figure 4, step 3 carries out early warning according to emergency to common anomalous ecg symptom using fuzzy algorithmic approach
Process are as follows: by sorting out anomalous ecg type, be classified into urgent, moderate is urgent, urgent three obscuring components of severe carry out it is pre-
It is alert；It is common to be divided into electrocardio stop fighting XT, twist mode ventricular tachycardia NS, ventricular fibrillation SC, third degree A-V block ZZ and myocardial infarction GS etc.
Symptom.
Such as the XT that fights is stopped for electrocardio, typical cardiac electrical figure shows as only seeing a series of sinus property P waves or room P wave and loseing
Any cross-connecting area or room escape, in the long pause for being longer than 2.7s (the longest cardiac cycle for being approximately equivalent to ventricular escape rhythm)
In have not seen any cross-connecting area and room property QRS wave.Using the value range of each characteristic component of electrocardiogram (ECG) data as the opinion of fuzzy set
Domain, wherein the linguistic variable meaning in fuzzy set is as follows:
NB- negative big (Negative Big), NS- bear small (Negative Small), and ZE- zero (Zero), PS- is just small
(Positive Small), PB- honest (Positive Big)；
The number domain of P wave group is [- D, D], and D value is 0.3, and fuzzy set is { NB, NS, ZE, PS, PB }, degree of membership
Function is all made of Triangleshape grade of membership function Trimf；RR interphase, QRS complex, P wave, T wave, PR interphase, QT interphase, ST sections be
Zero.
Output electrocardio stop the fighting domain of XT is [- F, F], and F value is 1, and fuzzy set is { NB, ZE, PB }, subordinating degree function
It is all made of Triangleshape grade of membership function Trimf；
Sentence meaning are as follows:
Such as: IF P is PB and RR is ZE and QRS is ZE and T is ZE and PR is ZE
and QT is ZE and ST is ZE Then XT is PB；Even P wave is more, RR interphase, QRS complex, P wave, T wave, PR
Interphase, QT interphase, ST sections of appearance are fewer, then electrocardio stops fighting more urgent.
For twist mode ventricular tachycardia NS, heart rate>160 (RR<280ms), QT extension>0.50S.Then the time domain of RR interphase is
[- E, E], E value are 0.3 (s), and fuzzy set is { NB, NS, ZE, PS, PB }, and subordinating degree function is all made of triangle degree of membership
Function Trimf；QT time interval domain is [- F, F], and F value is 1 (S), and the domain of output twist mode ventricular tachycardia NS is [- G, G], G
Value is 1, and fuzzy set is { NB, ZE, PB }, and subordinating degree function is all made of Triangleshape grade of membership function Trimf.
It is 250~500 times/min for ventricular fibrillation SC heart rate.Its judgement is relatively simple, it is only necessary to be divided into 250-350 times/min
(urgent), 350-450 times/min (moderate is urgent), 450-500 times/min (severe is urgent) section determines.
For third degree A-V block ZZ, heart rate < 45 beat/min, atrium is respectively exciting with ventricle, mutually incoherent, has been in
Full property chamber separation, QRS wave < 0.11S, then the time domain of QRS complex is [- H, H], and H value is 0.2 (s), and fuzzy set is
{ NB, NS, ZE, PS, PB }, subordinating degree function are all made of Gaussian subordinating degree function Gaussmf；
For myocardial infarction GS, T wave height is sharp, can reach 1.2mv-1.5mv, ventricular activation time on normal adult's chest leads
Extend 50ms, ST section are raised > 0.1mV, then the domain of T wave be [- I, I], I value be 2 (mv), fuzzy set for NB, NS, ZE,
PS, PB }, subordinating degree function is all made of bilateral Gaussian subordinating degree function Gauss2mf；The time domain of QR interphase is [- J, J],
J value is 0.05 (s), and fuzzy set is { NB, NS, ZE, PS, PB }, and subordinating degree function is all made of bilateral Gaussian degree of membership letter
Number Gauss2mf；ST sections of domain is [- K, K], and K value is 1 (mv), and fuzzy set is { NB, NS, ZE, PS, PB }, degree of membership
Function is all made of bilateral Gaussian subordinating degree function Gauss2mf.The domain for exporting myocardial infarction GS is [- L, L], and L value is 1,
Fuzzy set is { NB, ZE, PB }, and subordinating degree function uses bilateral Gaussian subordinating degree function Gauss2mf.
In addition to urgent situation, other slight arrhythmia cordis can also use fuzzy judgement, such as ventricular premature beat SZB to do sth. in advance
Occur QRS complex, before without and its relevant dystopy P wave；The roomy deformity of QRS complex, time are typically greater than or equal to 0.12
Second, and generally entailing increasing for R wave amplitude, T wave direction is contrary with the main wave of QRS complex.Therefore, the width of QRS complex is selected
Degree, RR interphase, T wave direction are as characteristic parameter, if the width of QRS wave is bigger, RR interphase is smaller, and T wave direction and QRS complex
Main wave is contrary, then is more likely to be ventricular premature beat.Using the value range of each characteristic component of electrocardiogram (ECG) data as fuzzy set
Domain, wherein the linguistic variable meaning in fuzzy set is as follows:
NB- negative big (Negative Big), NS- bear small (Negative Small), and ZE- zero (Zero), PS- is just small
(Positive Small), PB- honest (Positive Big)；
The width domain of QRS complex is [- D, D], and D value is 0.3, and fuzzy set is { NB, NS, ZE, PS, PB }, is subordinate to
Degree function is all made of Triangleshape grade of membership function Trimf；
The domain of RR interphase is [- E, E], and E value is 0.5, and fuzzy set is { NB, NS, ZE, PS, PB }, subordinating degree function
It is all made of Triangleshape grade of membership function Trimf；
T wave direction is primarily to see whether it consistent with the main wave direction of QRS complex, is indicated with two-value amount, even T wave with
The main wave direction of QRS complex is identical, takes 0, then takes 1 on the contrary；
The domain for exporting ventricular premature beat SZB is [- F, F], and F value is 1, and fuzzy set is { NB, ZE, PB }, degree of membership letter
Number is all made of Triangleshape grade of membership function Trimf；
Sentence meaning are as follows:
Such as: IF QRS is PB and RR is NB and QRS=0 (1) and T=1 (0) Then SZB is
PB；
The width of even QRS wave is bigger, RR interphase is smaller, and T wave direction is contrary with the main wave of QRS complex, then room property
Premature beat symptom is urgent.
2, doctor's client
Doctor's client is made of doctor's client software management, and above-mentioned warning information is by data processing module intelligence metaplasia
At, and it is shown in doctor's client application shown in fig. 5 interface, and after opening checks first aid information (as shown in Figure 5), first aid letter
Breath includes the detailed patient's condition of patient and diagnosis and treatment information, it is especially desirable to, it is noted that giving in Fig. 5 pre- in first aid information
An alert column provides early warning after the information of patient is collected processing for the judgement of doctor, when patient is diagnosed as ACVD and 120
After in emergency tender, in the case that condition is more urgent, exemplary early warning is gone out given in first aid information.
Doctor's client is the client software that remote emergency and health control can be run on intelligent terminal, including first aid
Management system and health management system arranged.Incident management system includes first aid electronic health record, rescue record, real-time monitoring etc., completely
Record the entire rescue process of patient.It is health management system arranged individual physiological index to be preserved automatically, in real time, it is convenient for
Observation and follow-up.System can be entered in conduit room, on ambulance by tablet computer, smart phone etc. to the operation of system
It is operated in equipment, it is simple and convenient.
3, network server
Network server mainly includes login interface, function of tonic chord interface, and exploitation environment is Visual studio 2010,
Dynamic web page interactive mode uses asp.net technology；Program uses the scripting language combination Ajax technology asynchronous refresh page,
ActiveX control is embedded in connect doctor's client, executes corresponding control instruction.Backstage is connected using the ADO.net based on C#
Database is connect, parameter information is read.Finally by Internet information service management device binding local IP address by website orientation to
Doctor's client WEB webpage.
The doctor of intelligent terminal browser address field input management information system for hospitals website IP address log in and
Browse webpage.Network server carries out authentication to it according to the information about doctor that database server stores, if do not led to
Verifying is crossed, then shows that authentication fails, doctor logs in again after need to checking account and password；If being cured by authentication
Raw can choose executes control instruction operation or inquiry monitoring operation.Control instruction operation is executed, then there is remote medical consultation with specialists, refer to
Lead rescue, real-time monitoring and instruction of changing the place of examination in time.And the doctor of inquiry monitoring operation is executed, it can check the personal letter of patient
Specific location locating for breath, treatment state, first aid information and patient.
4, database server
This system uses SQL Server2000 database, and 2000 database of SQL Server is that Microsoft releases
One data base management system, it is easy to use, has the characteristics that good scalability and integrated level are high, can make full use of
The advantage of Windows NT, it supports storing process, ODBC, ADO technology, and has autonomous sql like language, provides for developer
One good data management platform.It accesses database and uses ADO technology.ADO is that Microsoft is newest and most powerful
Data access paradigm OLE DB and design, be an application program layer interface easy to use.ADO technology is application program visit
It asks that database provides a simple, light and high performance interface, the least number of plies is used between front end and data source, it
Easy to use, speed is fast, memory less expenditure.
Database server is connected with network server, and when patient sees a doctor for the first time, database server is by its basic condition
It is transmitted to management information system for hospitals；It sees a doctor for the first time if patient is non-, doctor's client can log in management information system for hospitals inquiry
The related data of patient allows rescue personnel quickly to understand the basic condition and passing medical history of patient.
The data information that this system is related to includes following six class: personal patient information, doctor's essential information, medical monitoring number
According to the setting of, alarm parameter, idagnostic logout and GPS positioning.
(1) personal patient information includes the first aid ID of patient, name, gender, age etc..
(2) doctor's essential information includes doctor's number, name, department.
(3) medical monitoring data include ACS serum cardiac troponin I, ECG ST section, left ventricular ejection fraction, blood pressure, blood
Oxygen, blood glucose.
(4) alarm parameter setting includes EGC parameter type, the parameter value upper limit, parameter value lower limit, warning level.
(5) idagnostic logout includes patient ID, doctor ID, Diagnostic Time, diagnostic result.
(6) GPS positioning includes longitude, latitude.
Medical monitoring tables of data is as shown in table 3:
Table 3tb_Electrocardiodata
Alarm parameter setting is as shown in table 4:
Table 4tb_Alarmdata
| Column name | Data type | Description | Whether major key |
| AType | Int | EGC parameter type | It is |
| Aupper | Float(8) | Parameter upper limit | It is no |
| ALower | Float(8) | Parameter lower limit | It is no |
| ALevel | Int | Warning level | It is no |
Idagnostic logout is as shown in table 5:
Table 5tb_Diagnose Content
The mode advantage is: (1) the real-time 12 lead electrocardiogram remote transmission based on Internet of Things, helps base doctor couple
ACVD carly fruit drop reduces ACVD patient and postpones in basic hospital therapeutic time；(2) ACVD patient is directly entered by ambulance
Conduit room reduces patient and enters history-taking, physical examination, ECG examination and calling consultation of doctor time after emergency ward；(3) it is rescuing
Shield vehicle transhipment patient carries out conduit room's preparation simultaneously, reduces preoperative preparation time；(4) state of an illness, therapeutic scheme etc. are linked up
Work is completed in front of reaching intervention center, is improved patient and family members to the understanding of ACVD disease severity, is obtained and know together
The meaning time shortens；(5) classify method for early warning high safety and reliability, the Diagnostic Time of Medical Technologist is greatly saved, it not only can be with
Emergency cardiovascular care management mode is further improved, ACVD treatment delay is reduced, reduction hospitalization cost, improves prognosis, can be also doctor
Institute preferably utilizes platform of internet of things serve the people to provide Experience.
To sum up, above-mentioned first-aid centre's server system can realize that classification early warning scheme includes:
System is according to the concrete condition of the 12 collected patients of lead heart real time figure acquisition module, and GPS positioning is housed
The specific location of 120 ambulances of tracking system determines to send patient to collaboration and gives treatment to equipment in network, medical staff, distance
Matched network hospital.System decides whether that corresponding network hospital starting is notified to go forward side by side according to the urgency level of patient profiles
Enter conduit room's Preoperative Method.If the patient profiles on 120 ambulances are more urgent, Internet of Things information centre can pass through 4G wireless network
After patient's concrete condition is transmitted to primary network station hospital intelligent terminal, remote emergency software is run on intelligent terminal, intelligence is eventually
The primary network station hospital expert at end instructs point-of-care personnel to do corresponding treatment by remote medical consultation with specialists.Patient assessment is in two, three
When grade network hospital, primary network station hospital can check the real-time patient's condition of patient by 4G wireless network in intelligent terminal.If just
The conditions of patients examined in two, three-level network hospital deteriorates, and primary network station hospital can instruct two, three-level network hospital to arrange to suffer from
Person changes the place of examination in time.Network hospital before primary network station hospital arrangement patient referral in time with intended recipient patient is got in touch, and is built
If transporting green channel, and inform the concrete condition of the network hospital current patents of intended recipient patient, avoid secondary diagnosis, saves
About patient assessment's time.Medical patient, Internet of Things information centre will acquire the basic condition of patient for the first time, establish patient individual
Data.Non- patient medical for the first time, primary network station hospital intelligent terminal can investigate the personal information of patient, more so as to medical staff
The basic condition and passing medical history of patient are understood fastly.
In the description of this specification, reference term " one embodiment ", " some embodiments ", " illustrative examples ",
The description of " example ", " specific example " or " some examples " etc. means specific features described in conjunction with this embodiment or example, knot
Structure, material or feature are included at least one embodiment or example of the invention.In the present specification, to above-mentioned term
Schematic representation may not refer to the same embodiment or example.Moreover, specific features, structure, material or the spy of description
Point can be combined in any suitable manner in any one or more of the embodiments or examples.
Although having been given and describing the embodiment of the present invention, it will be understood by those skilled in the art that: not
A variety of change, modification, replacement and modification can be carried out to these embodiments in the case where being detached from the principle of the present invention and objective, this
The range of invention is defined by the claims and their equivalents.
Claims (8)
1. a kind of classification method for early warning for emergency cardiovascular care network system, which includes that doctor uses
Family terminal, first-aid centre's server system, 12 lead electrocardiogram acquisition modules, the 12 lead electrocardiogram acquisition module pass through indigo plant
The information of acquisition is sent to Medical flat computer by tooth, and sends first-aid centre's server system to by 4G wireless network, described
First-aid centre's server system completes the classification processing of information data, and is connected by network with clinician user terminal；
First-aid centre's server system includes database server, network server, data processing module, doctor client
End, the database server receive 4G wireless network transmission patient correlation acquisition data after, respectively with network server,
Data processing module realizes interconnection, and the network server, data processing module are connected with doctor's client；The network clothes
For business device for executing control instruction operation and executing inquiry monitoring operation, doctor can check the personal information of patient, treatment shape
Specific location locating for state, first aid information and patient；The database server is used for transmission, accesses and establish patient individual's letter
Breath, the setting of doctor's essential information, medical monitoring data, alarm parameter, idagnostic logout and GPS positioning information；The data processing
Module for electrocardiosignal denoising, extract denoising after signal temporal signatures and compression reconfiguration, to the Modulation recognition after reconstruct with
And early warning is carried out according to emergency to symptom；Characterized by comprising the following steps:
Step 1 is first filtered the low-frequency disturbance that is mingled in electrocardiosignal, electrode slice contact noise, using median filtering
Low frequency baseline drift interference is filtered out, formula is as follows:
Wherein, x (i) is electrocardiosignal sequence, and k is window width, and C is length of window, and y (i) is the signal after denoising, and x is original
Beginning signal, y are filtered signal,The respectively average value of original signal and filtered signal；
During being filtered out to high-frequency noise, wavelet coefficient is handled using soft-threshold function, formula is as follows:
Wherein, b, bt, T be respectively electrocardiosignal original wavelet coefficients, treated wavelet coefficient and given threshold value, sign ()
For function, SNR is signal-to-noise ratio, and MSE is root-mean-square error；
Step 2 extracts its temporal signatures to pretreated electrocardiosignal, generates training sample set, to electrocardiogram (ECG) data training sample
Base of this collection data characteristics carries out sparse coding, the number P of dictionary base is determined, at random from electrocardiogram (ECG) data training sample
It concentrates and extracts the initial dictionary D of training sample composition0∈RN×P, and the optimization of initial dictionary is carried out, random simulation generates some electrocardio
Data testing signals f ∈ RM×1, compression drop a dimension u=Φ f is carried out to the test signal, finally to electrocardiogram (ECG) data test signal f's
Reconstruct；D0It is sample for dictionary, R, the letter that M is the sampling number of each training sample, u is electrocardiogram (ECG) data to be obtained in sampling
Number, Φ is observing matrix；
Step 3 carries out classification processing to the electrocardiosignal after reconstruct using feature migration algorithm, obtains anomalous ecg symptom, and
Early warning is carried out according to emergency to anomalous ecg symptom using fuzzy algorithmic approach.
2. a kind of classification method for early warning for emergency cardiovascular care network system according to claim 1, which is characterized in that
The step 2 the following steps are included:
Step 2.1, it includes normal P wave, QRS complex, T wave, PR interphase, RR interphase, QT that electrocardiogram (ECG) data signal model is established in emulation
Interphase, ST section；It is E=[E that emulation, which generates a large amount of electrocardiogram (ECG) data signal composition training sample set,1,E1,…,E7]∈RM×W, wherein
W is total training sample number, and M is the sampling number of each training sample, and R is sample；
Step 2.2, it determines the number P of dictionary base, is concentrated from electrocardiogram (ECG) data training sample extract 72 training samples at random
Form initial dictionary D0∈RN×P, the objective function of initial dictionary optimization are as follows:Wherein A0It is training sample set E in dictionary D0On
Rarefaction representation matrix, λ are regularization parameter；
Step 2.3, observing matrix Φ ∈ RN×MThe random Gaussian that (N < < M) is formed using independent identically distributed Gaussian random variable
Matrix is realized, carries out compression dimensionality reduction u=Φ f to the test signal, obtains the low-dimensional compression observation signal u ∈ R of the test signalN ×1, complete the compression sampling of test signal；
Step 2.4, to the reconstruct of electrocardiogram (ECG) data test signal f, using based on l0The non-convex optimization of signal reconstruction under norm solves
Orthogonal matching pursuit algorithm in algorithm solves equation Φ Da=u, obtains rarefaction representation matrix a ∈ of the test signal under dictionary
RP×1；Anti- solution reconstruct electrocardiogram (ECG) data tests signal
3. a kind of classification method for early warning for emergency cardiovascular care network system according to claim 1, which is characterized in that
The step 3 carries out classification processing to the electrocardiosignal after reconstruct using feature migration algorithm, obtains the tool of anomalous ecg symptom
Body process are as follows:
Input: using the electrocardiogram (ECG) data test signal after reconstruct as source domain data set DS, tally set YS, by RR interphase,
QRS complex, P wave, T wave, PR interphase, QT interphase, ST sections be used as target domain data set DT, lower-dimensional subspace dimension k, parameter alpha,
β, γ, λ, benchmark classifier f, maximum number of iterations Tmax；Output: low dimension projective matrix W1；
Step 3.1: pass through the element of distributional difference matrix:Building
Edge distribution difference matrix L0, construct Laplacian Matrix M1, seek cross-cutting constraint factor di, while setting St、SbIt is 0；Wherein,
nS、nTRespectively former field number of samples and target domain number of samples, xi,xjRespectively sample to be tested, n are number；
Step 3.2:t=1；Wherein, t is the number of iterations；
Step 3.3: solving formulaOrObtain projection matrix W1；Wherein, X=DS∪DT,
C is classification number, LCFor condition distributional difference matrix, I is the matrix that element is 1, and K is iterative parameter, StTotal divergence square between class
Battle array, SbClass scatter matrix,For StNon-linear form；
Step 3.4: by DSAnd DTPass through matrix Z=W1 TX is mapped in k n-dimensional subspace n, obtains ZSAnd ZT, φ is diagonal function；
Step 3.5: in data set { ZS,YSOn training benchmark classifier f, and using the obtained classifier of training to ZTIt is marked
Note, obtains the tally set of target domain data
Step 3.6: using data set { XS,YSAndAccording to
Building condition distributional difference matrix LC；Simultaneously according to formulaWithS is solved respectivelybAnd St, when non-linear
ForWithWherein, NCFor the number of samples of C class, u(c)For the mean value of C classification, SwFor Scatter Matrix in class；
Step 3.7:t ← t+1；
Step 3.8: if t >=TmaxOrNo longer change, exports W1；Otherwise, 3.3 are gone to step.
4. a kind of classification method for early warning for emergency cardiovascular care network system according to claim 1, which is characterized in that
The step 3 carries out the process of early warning using fuzzy algorithmic approach to anomalous ecg symptom according to emergency are as follows: by sorting out the heart
Electricity stops fighting, the anomalous ecg type of twist mode ventricular tachycardia, ventricular fibrillation, third degree A-V block and myocardial infarction, be classified into it is urgent,
Moderate is urgent, urgent three obscuring components of severe carry out early warning；Stop fighting XT using Triangleshape grade of membership function Trimf for electrocardio
Training, for twist mode ventricular tachycardia NS, subordinating degree function is all made of Triangleshape grade of membership function Trimf training, for ventricular fibrillation SC letter
It is single to divide several sections judgements；For third degree A-V block ZZ, subordinating degree function is all made of Gaussian subordinating degree function
Gaussmf training, for myocardial infarction GS, subordinating degree function is using bilateral Gaussian subordinating degree function Gauss2mf training.
5. a kind of classification method for early warning for emergency cardiovascular care network system according to claim 1, which is characterized in that
First-aid centre's server system setting can obtain primary network station hospital, two grade network hospital in primary network station hospital, and three
Grade network hospital, 120 emergency tenders, high-risk Yi Fa group and the high-risk easy GPS positioning information for sending out individual present position, and level-one net
Network hospital has highest decision-making power；Primary network station hospital or two can be arranged in the 12 lead electrocardiogram acquisition module
In grade network hospital or three-level network hospital or 120 emergency platforms or high-risk Yi Fa group and high-risk easy hair individual.
6. a kind of classification method for early warning for emergency cardiovascular care network system according to claim 1, which is characterized in that
The clinician user terminal is patient monitor or medical PDA or mobile phone.
7. a kind of classification method for early warning for emergency cardiovascular care network system according to claim 1, which is characterized in that
The network server includes login interface, function of tonic chord interface, and exploitation environment is Visual studio 2010, dynamic web page
Interactive mode uses asp.net technology；Program uses the scripting language combination Ajax technology asynchronous refresh page, insertion
ActiveX control executes corresponding control instruction to connect doctor's client.
8. a kind of classification method for early warning for emergency cardiovascular care network system according to claim 1, which is characterized in that
The database server runs on win7 environment, using the 2000 data base administration system of SQL Server of Microsoft Corporation
System is used as developing instrument, and for the VC+6.0 of Microsoft as database front-end, CPU is AMD XP1800+, memory kingston
3G DDR, hard disk Dell 600G.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810289623.3A CN108335744B (en) | 2018-04-03 | 2018-04-03 | A kind of emergency cardiovascular care network system and its method for early warning of classifying |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810289623.3A CN108335744B (en) | 2018-04-03 | 2018-04-03 | A kind of emergency cardiovascular care network system and its method for early warning of classifying |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108335744A CN108335744A (en) | 2018-07-27 |
| CN108335744B true CN108335744B (en) | 2019-01-11 |
Family
ID=62932018
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810289623.3A Expired - Fee Related CN108335744B (en) | 2018-04-03 | 2018-04-03 | A kind of emergency cardiovascular care network system and its method for early warning of classifying |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108335744B (en) |
Families Citing this family (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110600126A (en) * | 2019-09-19 | 2019-12-20 | 江苏大学附属医院 | Auxiliary evaluation method for arterial calcified plaque image of lower limb of diabetic foot |
| CN112568911A (en) * | 2019-09-30 | 2021-03-30 | 深圳市理邦精密仪器股份有限公司 | Method and equipment for classifying electrocardiogram data and device with storage function |
| CN116933166A (en) * | 2019-11-28 | 2023-10-24 | 太原理工大学 | Cerebral apoplexy-oriented unbalanced data set classification method and system |
| CN113160949A (en) * | 2020-01-20 | 2021-07-23 | 深圳市理邦精密仪器股份有限公司 | Method and server for providing doctor suggestion and mobile electrocardiogram equipment |
| CN112652384A (en) * | 2020-01-20 | 2021-04-13 | 深圳市理邦精密仪器股份有限公司 | Alarm reminding method and equipment for electrocardiograph, electrocardiograph and storage device |
| CN111329468A (en) * | 2020-02-10 | 2020-06-26 | 郭洪军 | Real-time monitoring system and monitoring method for cardiology department |
| CN111180091A (en) * | 2020-03-06 | 2020-05-19 | 东北大学 | Monitoring system for intelligent medical community service |
| CN111835360B (en) * | 2020-07-14 | 2021-06-01 | 山东省人工智能研究院 | Sparse signal reconstruction method based on basis pursuit denoising |
| CN112116979B (en) * | 2020-08-11 | 2024-03-19 | 重庆华医康道科技有限公司 | Electronic prescription circulation security working method based on blockchain account distribution protocol |
| CN112487880B (en) * | 2020-11-13 | 2021-09-24 | 河北省科学院应用数学研究所 | Median filter, device and system based on maximum error criterion |
| CN116548981B (en) * | 2023-07-12 | 2023-09-08 | 四川省医学科学院·四川省人民医院 | Wireless transmission split type electrocardiograph monitoring system |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN201259720Y (en) * | 2008-06-02 | 2009-06-17 | 深圳市安鹏达科技有限公司 | Life pre-alarm device and life pre-alarm rescue system |
| CN103690240A (en) * | 2013-09-16 | 2014-04-02 | 上海华美络信息技术有限公司 | Medical system |
| CN107736887A (en) * | 2017-11-09 | 2018-02-27 | 湖北工业大学 | A kind of Multifunctional mobile electrocardiogram monitor system and method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10531806B2 (en) * | 2013-12-17 | 2020-01-14 | University Of Florida Research Foundation, Inc. | Brain state advisory system using calibrated metrics and optimal time-series decomposition |
-
2018
- 2018-04-03 CN CN201810289623.3A patent/CN108335744B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN201259720Y (en) * | 2008-06-02 | 2009-06-17 | 深圳市安鹏达科技有限公司 | Life pre-alarm device and life pre-alarm rescue system |
| CN103690240A (en) * | 2013-09-16 | 2014-04-02 | 上海华美络信息技术有限公司 | Medical system |
| CN107736887A (en) * | 2017-11-09 | 2018-02-27 | 湖北工业大学 | A kind of Multifunctional mobile electrocardiogram monitor system and method |
Non-Patent Citations (1)
| Title |
|---|
| 结合小波变换和中值滤波心电信号去噪算法研究;徐寒等;《自动化与仪表》;20121231;第46-49页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108335744A (en) | 2018-07-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108335744B (en) | A kind of emergency cardiovascular care network system and its method for early warning of classifying | |
| CN105162880B (en) | The aid decision-making method of the collaborative treatment of a kind of ACS compartmentalization network | |
| Kachuee et al. | Ecg heartbeat classification: A deep transferable representation | |
| Gogate et al. | Healthcare monitoring system based on wireless sensor network for cardiac patients | |
| CN107103182A (en) | A kind of heart disease Warning System and method based on deep learning algorithm | |
| Shaown et al. | IoT-based portable ECG monitoring system for smart healthcare | |
| Kwon et al. | Artificial intelligence for detecting electrolyte imbalance using electrocardiography | |
| Faust et al. | Automated arrhythmia detection based on RR intervals | |
| US20210353203A1 (en) | Diagnostics for detection of ischemic heart disease | |
| El Attaoui et al. | Wearable wireless sensors network for ECG telemonitoring using neural network for features extraction | |
| Rahman et al. | Internet of things based electrocardiogram monitoring system using machine learning algorithm | |
| CN106510687A (en) | Method and system for identifying abnormal electrocardiogram data | |
| CN110200619A (en) | It is a kind of to establish major cardiovascular disease risks Early-warning Model method and device | |
| Zou et al. | Automatic detection of congestive heart failure based on multiscale residual unet++: From centralized learning to federated learning | |
| Filist et al. | Classifier for the functional state of the respiratory system via descriptors determined by using multimodal technology | |
| CN110349641A (en) | A kind of health and fitness information inquiry unit convenient for operation | |
| Pimentel et al. | Human mental state monitoring in the wild: Are we better off with deeperneural networks or improved input features? | |
| Liu et al. | Real time arrhythmia monitoring and classification based on edge computing and dnn | |
| SC et al. | Telecardiology for effective healthcare services | |
| Işik et al. | Pulse oximeter based mobile biotelemetry application | |
| Patil et al. | Detection of Cardiac Abnormalities and Heart Disease Using Machine Learning Techniques | |
| Singh et al. | Remote Monitoring System of Heart Conditions for Elderly Persons with ECG Machine Using IOT Platform | |
| Hu et al. | Cloud-assisted home health monitoring system | |
| Bollepalli et al. | Dictionary-based monitoring of premature ventricular contractions: An ultra-low-cost point-of-care service | |
| Maity et al. | Prototype Design of An IoT Enabled Cost-Efficient Portable Heart-Health Data Acquisition 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20190111 Termination date: 20200403 |