CN106344005B - A kind of removable electrocardiogram monitoring system - Google Patents
A kind of removable electrocardiogram monitoring system Download PDFInfo
- Publication number
- CN106344005B CN106344005B CN201610973366.6A CN201610973366A CN106344005B CN 106344005 B CN106344005 B CN 106344005B CN 201610973366 A CN201610973366 A CN 201610973366A CN 106344005 B CN106344005 B CN 106344005B
- Authority
- CN
- China
- Prior art keywords
- signal
- cnn
- uwb
- layer
- ecg
- 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.)
- Active
Links
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/318—Heart-related electrical modalities, e.g. electrocardiography [ECG]
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/68—Arrangements of detecting, measuring or recording means, e.g. sensors, in relation to patient
- A61B5/6801—Arrangements of detecting, measuring or recording means, e.g. sensors, in relation to patient specially adapted to be attached to or worn on the body surface
- A61B5/683—Means for maintaining contact with the body
- A61B5/6832—Means for maintaining contact with the body using adhesives
- A61B5/6833—Adhesive patches
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7203—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/725—Details of waveform analysis using specific filters therefor, e.g. Kalman or adaptive filters
Abstract
The invention discloses a kind of removable electrocardiogram monitoring systems, not only acquire ECG signal, and acquisition pulse radio ultra-wideband (IR-UWB) radar signal, after ECG signal collected and IR-UWB radar signal are synchronized and are handled, feature extraction, integrated analysis and diagnostic classification are carried out using cascade CNN, obtain monitoring result output.Since the embodiment of the present invention introduces IR-UWB radar signal in monitoring, supplement as monitoring ECG signal, and pass through the synchronization to both signals, obtain the correlation between both signals, improve monitoring accuracy, feature extraction, integrated analysis and diagnostic classification are carried out since the embodiment of the present invention uses adaptable and stable performance cascade CNN, stable system performance is effectively ensured.
Description
Technical field
The present invention relates to the monitoring technology of human body physiological parameter, in particular to a kind of removable electrocardiogram monitoring system.
Background technique
Heart disease seriously threatens the life and health of the mankind, and morbidity and mortality all come first of various diseases and the heart
The sudden of popular name for increases treatment difficulty.Therefore, prevention and treatment heart disease just becomes the cardinal task that the people safeguard health.
Removable electrocardiogram (ECG) monitoring system provides good solution for effectively prevention heart disease.On the one hand, it conveniently takes
With the characteristics of allow this monitoring system to monitor the ECG signal of cardiomotility for a long time, so as to capture in time
Sporadic arrhythmia cordis and other sudden heart diseases.On the other hand, with the fast development of mobile Internet of Things and information technology,
Such ECG monitoring system can obtain good platform and technical support, help preferably to meet and carry out whenever and wherever possible
The demand of heart health.
Currently, can be realized based on internet of things system structure when realizing ECG monitoring system, be based on Internet of Things net system knot
The ECG monitoring system of structure is representational three-decker, i.e. sensing layer, network layer and application layer.Wherein, sensing layer is mainly born
Electrocardiographicdata data acquisition is blamed, network layer is responsible for collected ECG data being transmitted to data center, and application layer has been responsible for
At the analytical calculation and problem early warning of electrocardiogram (ECG) data.When acquiring ECG data, various wearable Medical Devices can be used
It is connected in sensing layer by wireless data communication technology, forms a wireless sensor network.Network side can be by existing
Internet or backbone communication network undertake.Application layer is then to be detected by the various algorithms of application and analyze arrhythmia cordis.
Analyzing and diagnosing clearly for arrhythmia cordis is the core function of entire ECG monitoring system.
The analyzing and diagnosing that arrhythmia cordis is carried out based on ECG signal is a kind of typical classification application, existing arrhythmia cordis
Classification method mainly includes the mode of frequency analysis, support vector machines, wavelet transformation and expert's mixing.In short, these sides
Method all relies on greatly the feature that certain keys are extracted from ECG signal waveform.But to suffer from feature relevant for these modes
The limitation of threshold value, and the threshold value of different data sets is usually different, therefore, the performance of these methods not only will receive various make an uproar
The interference of sound and decline, also suffer from data set variation influence and it is unstable.
In order to cope with the above problem, the variable ECG signal of monitoring is classified more accurately, beginning to use has height fault-tolerant
The neural network of degree ability and robustness.For example, neural network can be trained to as a heartbeat identification detector, Yi Zhongxin
Not normal classifier or a feature extractor are restrained, for extracting the human body physiological parameter not being manually set before classification
Feature.There is the automatic ability for obtaining feature just because of neural network, ECG monitoring system neural network based can dispose
In mobile terminal and cloud, by periodically learning new feature from the cloud ECG of specific user record, so as to preferably
Suitable for specific user, and help to improve the accuracy of classification.Up to the present, many has been had already appeared about in the dynamic heart
The research of arrhythmia classification is carried out in electrograph monitoring system to specific user.
Specifically, convolutional neural networks (CNN) are one of the research hotspots of deep learning in recent years, there is structure letter
List, training parameter are few and adapt to the features such as strong, are widely used in the fields such as pattern-recognition and image procossing, achieve well
Effect.ECG signal analysis is a kind of application of image steganalysis, therefore also has researcher that CNN is applied to this, and
Illustrate that CNN ECG signal two-dimensional for a peacekeeping is all applicable.A kind of lead convolutional neural networks are disclosed at present
(LCNN), eight leads ECG signal recorded are converted into a two-dimensional matrix.Wherein, the sliding scale of convolution kernel is filtered
Wave device cannot be recorded shared rule by the ECG signal of different leads and limit, and this method not only demonstrates CNN in one-dimensional signal
In availability, but also adapt in different resolution ratio, although the accuracy of ECG classification can be improved in this method, so
And the application of this method is limited, and is required to the multi-lead input of ECG signal,
While the sorting algorithm research of ECG signal constantly promotes, another improves the striving direction of system performance just
It is the data collection terminal in ECG signal.In fact, why ECG monitoring system can reduce the cardiopathic death rate, it is basic former
Because being that it can record ECG signal to long-term dynamics.In this case, even some rare arrhythmia cordis phenomenons
Also it can be caught in, so as to treat in time, save sufferer.But ECG signal record inevitably will receive due to people
The influence for the motion artifacts that body muscular movement generates causes sorting algorithm accuracy to decline.Although some movements have been proposed
Artifact inhibition algorithm, such as adaptive filter method and blind source separation method, but it is still highly difficult for completely removing motion artifacts
's.
To sum up, it based on the inaccuracy and the subsequent inaccuracy in classification ECG signal when monitoring ECG signal, causes
Current ECG signal monitoring system is inaccurate when recording to obtain monitoring result based on the ECG signal monitored, causes to people
The erroneous judgement of the arrhythmia cordis phenomenon of body.
Summary of the invention
In view of this, the embodiment of the invention provides a kind of removable electrocardiogram monitoring system, which enables to base
It is accurate in the monitoring result that the ECG signal monitored records.
The embodiment of the present invention also provides a kind of removable electrocardiogram monitoring method, and this method is enabled to based on being monitored
The monitoring result that ECG signal records is accurate.
As can be seen from the above scheme, the present invention implements are as follows:
A kind of removable electrocardiogram ECG monitoring system, comprising: acquisition module and cascade CNN module, wherein
Acquisition module, for monitoring impulse radio ultra-wideband IR-UWB radar signal, after synchronizing and integrating processing
Obtain IR-UWB heartbeat signal；ECG signal is monitored, obtains ECG signal data after synchronizing processing and unbalanced data processing；
Concatenated convolutional neural network CNN module, for from the received IR-UWB heartbeat signal of acquisition module and it is integrated from
After ECG signal data after reason carries out feature extraction, integrated and diagnostic classification, monitoring result is obtained.
A kind of ECG monitoring method, comprising:
IR-UWB radar signal is monitored, synchronize and obtains IR-UWB heartbeat signal after integrating processing；
ECG signal is monitored, obtains ECG signal data after synchronizing processing and unbalanced data processing；
To from the received IR-UWB heartbeat signal of acquisition module and integrated treated that ECG signal data carries out feature pumping
Take, integrate and diagnostic classification after, obtain monitoring result.
From above scheme as can be seen that the embodiment of the present invention not only acquires ECG signal, but also acquisition pulse radio ultra-wide
Band (IR-UWB) radar signal, after ECG signal collected and IR-UWB radar signal are synchronized and handled, using cascade
CNN carry out feature extraction, integrated analysis and diagnostic classification, obtain monitoring result output.Since the embodiment of the present invention is monitoring
When introduce IR-UWB radar signal, as the supplement of monitoring ECG signal, and by the synchronization to both signals, obtain this
Correlation between two kinds of signals improves monitoring accuracy, stablizes since the embodiment of the present invention uses adaptable and performance
Cascade CNN carry out feature extraction, integrated analysis and diagnostic classification, stable system performance is effectively ensured.Therefore, using the present invention
The system and method that embodiment provides makes the monitoring result recorded based on the ECG signal monitored accurate.
Detailed description of the invention
Fig. 1 is a kind of removable electrocardiogram monitoring system structure diagram provided in an embodiment of the present invention；
Fig. 2 is that the structure that synchronous submodule provided in an embodiment of the present invention carries out integrated processing to IR-UWB radar signal is shown
It is intended to；
Fig. 3 is the structural schematic diagram of the first CNN or the 2nd CNN provided in an embodiment of the present invention；
Fig. 4 is the structural schematic diagram of the 3rd CNN provided in an embodiment of the present invention；
Fig. 5 is a kind of removable electrocardiogram monitoring method flow diagram provided in an embodiment of the present invention；
Fig. 6 is the specific example structure chart of the first CNN provided in an embodiment of the present invention；
Fig. 7 is the specific example structure chart of the 2nd CNN provided in an embodiment of the present invention；
Fig. 8 is the specific example structure chart of the 3rd CNN provided in an embodiment of the present invention.
Specific embodiment
To make the objectives, technical solutions, and advantages of the present invention more comprehensible, right hereinafter, referring to the drawings and the embodiments,
The present invention is described in further detail.
The embodiment of the present invention is in order to accurately monitor arrhythmia cordis under motion state, by IR-UWB radar signal
Monitoring technology be introduced into ECG monitoring system provided by the invention, as auxiliary monitoring tool.IR-UWB radar not only provides
A kind of measurement of contactless IR-UWB radar signal, while also avoiding electrical interference.The clinical application packet of IR-UWB radar
Include the measurement to heart rate, respiratory rate and blood pressure etc..Using the embodiment of the present invention, compared with the background art, mainly from two angles
To enhance system under motion state to the robustness of arrhythmia cordis monitoring function.First, IR-UWB radar signal is introduced, as
To effective supplement of ECG signal record, also, by being synchronized to both signals and integrated analysis, can also excavate
Potentially relevant property between both data, it is more strong to be provided for the diagnostic classification of arrhythmia cordis in dynamic monitoring system
Information foundation, improve the accuracy of monitoring.Second, using adaptable and stable performance cascade CNN carry out feature extraction,
Integrated analysis and diagnostic classification, help to maintain stable system performance.
Fig. 1 be ECG monitoring system structure diagram provided in an embodiment of the present invention, as shown, include acquisition module and
Cascade CNN module, wherein
Acquisition module obtains IR-UWB heartbeat letter after synchronizing and integrating processing for monitoring IR-UWB radar signal
Number；ECG signal is monitored, obtains ECG signal data after synchronizing processing and unbalanced data processing；
CNN module is cascaded, for from the received IR-UWB heartbeat signal of acquisition module and integrated treated ECG signal
After data carry out feature extraction, integrated and diagnostic classification, monitoring result is obtained.
In this configuration, the acquisition module further includes ECG signal acquisition submodule, for monitoring ECG signal.ECG letter
The biological electrical feature for number carrying heartbeat, such as ventricular depolarization and polarized interval again.ECG signal acquisition submodule can be based on
ECG sensor chip BMD101 and mating bluetooth submodule are built.Although BMD101 chip can only obtain ECG signal one leads
Connection, but it is enough for the diagnostic classification of subsequent normal/abnormal heartbeat.
In this configuration, the acquisition module further includes IR-UWB radar signal acquisition submodule, for monitoring IR-UWB
Radar signal.NVA-R661 radar module can be used to construct IR-UWB radar acquisition submodule, due to the transmitting function of radar
Rate is 1/10th also lower than wireless network transmission power, so injury will not be brought to human body.
Although ECG signal and the IR-UWB radar signal from radar have different features, they can be total to
With description heartbeat parameter, it can complement each other, especially be to have very much in the case where wherein a kind of data are heavily disturbed
Meaning.Subsequent cascade CNN resume module for convenience has also carried out the processing of synchronous and unbalanced data.In general, ECG is adopted
The sample rate for collecting submodule is 512 hertz (Hz), and when radar covering radius is set as 2 meters, the acquisition of IR-UWB radar signal
The sample rate of submodule only has 38Hz.Since the sample rate of both signals is different, so that treated ECG signal and IR-UWB
The length of heartbeat signal is also different, to be respectively both signals in cascade CNN module progress integrated analysis, extraction feature
Different CNN is set.
Since ECG signal and IR-UWB radar signal are acquired by different acquisition channels, when acquisition originates
Between and transmission time may all can be different.Even so, even if BMD101 chip and NVA-R661 radar module open simultaneously
Dynamic, respectively required time is also different before their life's works.Therefore, more accurately collected in order to subsequent
At analysis, it is necessary to be synchronized to both signals, that is, time calibration.In embodiments of the present invention, when having selected reception
Between stamp be used as the standard of time calibration.
In this configuration, acquisition module further includes integrated processing submodule, for the IR-UWB radar signal to monitoring into
The integrated processing of row.
Fig. 2 is that the structure that synchronous submodule provided in an embodiment of the present invention carries out integrated processing to IR-UWB radar signal is shown
It is intended to, comprising: remove direct current component, band-pass filter unit, principal component analysis (PCA) removal noise signal unit, extract mainly
Signal element and integrated empirical mode decomposition unit, wherein
Direct current component is removed, for removing the DC component in IR-UWB radar signal；
Band-pass filter unit, for IR-UWB radar signal to be carried out bandpass filtering；
PCA removes noise signal unit, for removing the clutter in IR-UWB radar signal；
Main signal unit is extracted, for extracting the IR-UWB heartbeat signal in IR-UWB radar signal；
Integrated empirical mode decomposition unit, for obtained IR-UWB heartbeat signal rule of thumb Mode Decomposition will to be extracted,
Breath signal therein is removed, the IR-UWB heartbeat signal of needs is finally obtained.
Since IR-UWB radar has high spatial resolution and strong penetrability, this is highly suitable for it at monitoring
Thoracic cavity under motion state rises and falls.In the IR-UWB radar signal obtained by monitoring, in addition to the IR-UWB heartbeat letter of needs
Extra needs wherein also mixing DC component, noise signal and breath signal etc. using following specific signal processing methods
Therefrom obtain IR-UWB heartbeat signal.
In the IR-UWB radar signal matrix that monitoring obtains, the τ when time of every a line is referred to as fast, the time of each column
T when being slow.Every a line in matrix is the sampling of signal waveform, and each column then represent different positions of the echo apart from radar
It sets.Monitor obtained IR-UWB radar signal and contain DC component, be static constant, will not with it is slow when t change and change
Become.Therefore, it can ask row average by every row to signal matrix, then subtract row averagely with every row of original signal, thus
Remove the DC component in IR-UWB radar signal matrix.
The frequency range of signal can be received by the setting control to parameter in NVA-R661 radar module.Known
In the case where parameter setting, it is known that frequency range, then the embodiment of the present invention can using a band-pass filter unit
The target IR-UWB radar signal needed is obtained with further.
PCA mode is commonly used for extracting the main feature ingredient of data.And noise signal is since static background environment is anti-
Caused by having penetrated radar signal, so back wave of the noise signal in same position, i.e., change very in the same row of radar matrix
It is small.Therefore, the energy of PCA noise signal is typically greater than the energy of the echo-signal caused by thoracic cavity rises and falls.Then, pass through surprise
Different value is decomposed (SVD) formula and is decomposed to signal matrix:
R=USVT(formula 1)
Wherein, S is made of the non-negative singular value of R, the element in S, in addition to the element on diagonal line is all 0, diagonally
Line element is then the non-negative singular value arranged according to descending, can be indicated are as follows:
S1,1> S2,2> ... > SN, N
These values have corresponded to the size of energy, and therefore, the energy of IR-UWB radar signal is mainly distributed on the several of front
In singular value.The quantity of energy ratio specific for one, corresponding singular value also determines that.If we are these singular values
It is set as 0, then the energy of corresponding proportion is just removed.Then, it is rebuild after SVD, most of energy of noise signal
It is just removed, the IR-UWB radar signal matrix of reconstruct will have higher signal-to-noise ratio.
In IR-UWB radar signal matrix at this moment, all column correspond to the energy that different line positions are set.Since environment is
Static, corresponding that in fluctuating thoracic cavity possesses maximum variance once arranging.The embodiment of the present invention calculates IR-UWB radar letter as a result,
The variance that number matrix respectively arranges finds that maximum column of variance, has just obtained IR-UWB heartbeat signal.
After IR-UWB heartbeat signal column has been determined, the mixing from breath signal and IR-UWB heartbeat signal is needed
IR-UWB heartbeat signal is separated in signal.Empirical mode decomposition (EMD) is according to signal self-characteristic, by determining IR-
UWB heartbeat signal is decomposed into a series of intrinsic mode functions (IMF), and each IMF embodies different frequency scale in signal
Shake characteristic.Integrated empirical mode decomposition (EEMD) is the improved method based on EMD.EEMD is in determining IR-UWB heartbeat signal
Random white noise is added, then repeatedly signal is handled with EMD, averages to obtained IMF is all decomposed as final
IMF, the i.e. basic function of determining IR-UWB heartbeat signal.Obtained IMF component is arranged from high frequency to low frequency sequence, is sorted low
IMF component represent concussion mode faster, the IMF component that sorts high represents slower concussion mode.To each IMF, meter
Calculate its gross energy (ETotal) ETotal of entire frequency range and in heartbeat frequency range ([0.8Hz, 2.5Hz]) energy
(EHeart).According to the energy ratio of EHeart and ETotal, the IMF with ceiling capacity ratio is selected to reconstruct IR-UWB heartbeat letter
Number.
In this configuration, acquisition module further includes unbalanced data processing submodule, is also used to carry out ECG signal uneven
It weighs after data processing, increases the abnormal heartbeats number in ECG signal.Herein, believe after unbalanced data processing using to ECG
Number carry out over-sampling mode and addition white noise mode.
In general, normal heartbeat often quantitatively substantially exceeds abnormal heartbeats, has thereby resulted in ECG signal
Imbalance, and further cause major class accuracy also above group accuracy.In order to handle this unbalanced signal collection,
Two methods are proposed for inhibiting influence of the major class to whole classification accuracy.A kind of method is increased in signal collection side
Group reduces major class, and another method is then to add weight to classifier.Using CNN, over-sampling mode ratio adds
Power classifier methods perform better than.Therefore, the embodiment of the present invention has used over-sampling mode.For the periodic signal of ECG signal,
A kind of over-sampling mode is to increase number of samples by moving the starting point of sampling in one cycle.In addition, in ECG signal
Middle addition noise can also be improved CNN classification accuracy.These newly-increased noises are similar to baseline drift and Hz noise, meet
The principle that can be removed.The embodiment of the present invention also uses both methods to increase the sample number of abnormal heartbeats.
In this configuration, cascade CNN module is responsible for integrated treated ECG signal data and the IR-UWB heartbeat of integrated analysis
Signal, to diagnose normal heartbeat/abnormal heartbeats.The structure of cascade CNN module mainly has two layers, and first layer includes two differences
CNN, i.e. the first CNN, for IR-UWB heartbeat signal carry out feature extraction, extract feature.And the 2nd CNN, for pair
It is integrated that treated that ECG signal data carries out feature extraction.The embodiment of the present invention uses this set, as previously mentioned, be due to
Obtained ECG signal data and the IR-UWB heartbeat data obtained from radar have different sample rate and data structure.It is special
Not, feature here is not artificial pre-designed, such as wave amplitude, area, mean value, but is excavated by CNN oneself
Data in hide further feature, also be exactly this feature so that CNN is at present best in the performance of field of image processing
's.Moreover, because ECG signal data and IR-UWB heartbeat signal be all it is one-dimensional, the first CNN and the 2nd CNN also accordingly divide
Feature extraction Cai Yong not carried out using one-dimensional convolution kernel.The second layer for cascading CNN is the 3rd CNN, it is responsible for the first CNN
And the 2nd the feature that extracts of CNN carry out classification mark, label has normal heartbeat and two kinds of abnormal heartbeats.It can be instructed by study
Practice to promote the accuracy of the 3rd CNN classification.After training, can real time execution, Dynamic Display goes out monitoring result.?
One layer between the second layer, the embodiment of the present invention is arranged feature and integrates submodule, is used to take out the first CNN and the 2nd CNN
The two kinds of features taken out are integrated, and core here is Integrated Strategy.A two dimensional character is features defined after integrated
Collection, is input to the 3rd CNN, therefore the convolution kernel of the 3rd CNN is two-dimensional.The classification results of 3rd CNN may also pass through assessment
It feeds back feature and integrates submodule, to adjust Integrated Strategy, optimization system performance.
Fig. 3 is the structural schematic diagram of the first CNN or the 2nd CNN provided in an embodiment of the present invention, due to the first CNN and second
The structure of CNN is identical, so directly explanation, comprising: signals layer, convolutional layer 1, pond layer 1, convolutional layer 2, pond layer 2, convolutional layer
3 and pond layer 3, processing is all one-dimensional case.The structure shares 3 convolutional layers, a then pond after each convolutional layer
Layer.Input data is all that one-dimensional, all convolution kernel is also all one-dimensional.In the forward propagation process, per one-dimensional convolutional layer
Calculation formula it is as follows:
Wherein,It is l layers of kth characteristic pattern,It is the biasing of kth characteristic pattern in l layers,It is l-
1 layer of i-th characteristic pattern,It is the convolution mapped from l-1 layers of all features to l layers of k-th of characteristic pattern
Core, N are the total characteristic numbers in l-1 layers.Symbol conv indicates that Vector convolution, f () indicate excitation function.
Can have multiple convolution kernels, each convolution kernel corresponds to an one-dimensional filter, to extract it is a kind of without
The feature of design generates a characteristic pattern.In fact, due to IR-UWB heartbeat signal scale unlike ECG signal data so
Greatly, the neuron number in each convolutional layer of the first CNN and the 2nd CNN, convolution kernel size can not be identical, but they
Output feature want to be input in the second layer jointly, so, they need one-dimensional characteristic of the output phase with scale.
Fig. 4 be the 3rd CNN provided in an embodiment of the present invention structural schematic diagram, including input layer, convolutional layer, pond layer,
Hidden layer and output layer form the 3rd CNN's specifically, the feature of extraction is exported and integrated by the first CNN and the 2nd CNN
Two-dimensional matrix is input to input layer, after the convolutional layer that two-dimensional convolution core is possessed by one, connects a pond layer, then through complete
Connection enters a hidden layer, is finally the result that output layer provides classification.In fact, subsequent two layers is exactly one classical
Multi-layer perception (MLP) (MLP).Herein, come integrated study two by using two-dimensional convolution core to be given birth to by the first CNN and the 2nd CNN
At one-dimensional characteristic between potentially relevant property, help the biomedical record of multi-platform channel more accurately to assign to difference
In classification.Even if sometimes ECG signal data record or IR-UWB heartbeat signal have one kind destroyed, final output knot
Fruit still may rely on both one-dimensional characteristics and the potentially relevant property between them, to keep preferable stability.
Particularly, Chinese cardiovascular disease database (CCDD) provides some available ECG signal resources, and the present invention is real
The pre-training for having supervision can be carried out to the 2nd CNN handled for ECG signal by applying example.The embodiment of the present invention has used 150,000
ECG signal record in CCDD, the time of each record second are differed from 10 seconds to 20, the signal comprising 12 leads.By
After data nonbalance processing, normal heartbeat and abnormal heartbeats number are roughly equal, so that it may use as training dataset.
In this configuration, the extraction feature of the first CNN and the 2nd CNN first passes through feature set before being input to the 3rd CNN
At so that the 3rd CNN can not only utilize IR-UWB heartbeat signal and the respective feature of ECG signal data, additionally it is possible to excavate
Potentially relevant property between the two.In embodiments of the present invention, this correlation is exactly to improve holistic diagnosis classification accuracy
Critical support and the embodiment of the present invention introduce IR-UWB heartbeat signal to supplement the main purpose of ECG signal data disadvantage.When
So, CNN itself has had preferable integrated analysis ability, if the embodiment of the present invention can be further improved its input signal
Integrated advantage, then overall performance will further get a promotion.Therefore, increase feature integrate submodule purport should be as
What is integrated IR-UWB heartbeat signal and ECG signal data are appropriate, the complementarity that both can preferably embody and
Correlation.
In CNN module increase feature integrate submodule mainly consider of both factor.First, IR-UWB heartbeat signal
Processing and feature extraction are carried out in channel independent with ECG signal data, obtained result value range may phase
Is difference bigger, and this gap can produce bigger effect result? need to be standardized? second, how accurately to sentence
Complementary presence and it is used between disconnected two kinds of signals? if IR-UWB heartbeat signal and ECG signal data are one of
Impaired, integrated result is inevitable at this time takes into account potentially relevant property between the two based on undamaged one kind, this is exactly of the invention
Embodiment expects the case where performance can most get a promotion.How is this handled but if two kinds of signals are all impaired? if two kinds of signals
All very well, negative correlation but is presented, and how this is handled?
Based on above-mentioned analysis, the feature that the embodiment of the present invention proposes is integrated submodule and is integrated using weighting, is expressed as follows:
Y=[a (1- θ1)x1, b (1- θ2)x2]
Wherein, y represents integrated 2D signal, x1Represent IR-UWB heartbeat signal, x2ECG signal is represented, a represents heartbeat
The normalization factor of signal, b represent the normalization factor of ECG signal, θ1∈ [0,1] represents the impaired system of IR-UWB heartbeat signal
Number, θ2∈ [0,1] represents the impaired coefficient of ECG signal, therefore, (1- θ1) represent the intact degree of IR-UWB heartbeat signal, (1-
θ2) represent the intact degree of ECG signal.As it can be seen that it is flat to can control the size between two kinds of signal dimensions by normalization factor
Weighing apparatus, by being damaged a kind of contribution specific gravity of the adjustable signal of coefficient in integrated signal.In general, normalization factor can root
It is set according to truthful data situation, impaired coefficient can be set according to the working condition of data acquisition platform.Obtaining performance evaluation
After feedback, also adjustable above-mentioned parameter.
Fig. 5 is a kind of removable electrocardiogram monitoring method flow diagram provided in an embodiment of the present invention, the specific steps are that:
Step 501, monitoring IR-UWB radar signal synchronize and obtain IR-UWB heartbeat signal after integrating processing；
Step 502, monitoring ECG signal obtain ECG signal data after synchronizing processing and unbalanced data processing；
Step 503, to from the received IR-UWB heartbeat signal of acquisition module and integrated treated that ECG signal data carries out
After feature extraction, integrated and diagnostic classification, monitoring result is obtained.
A specific example is lifted the embodiment of the present invention is described in detail.
IR-UWB radar signal and ECG signal are respectively by radar module NVA-R661 and ECG module BMD101 sensor core
Piece acquisition.Subject person is sitting in front of IR-UWB radar, is sticked electrode slice in wrist and is connected with ECG sensor.Share 20 volunteers
Take part in data acquisition.Unbalanced data is handled using oversampler method.
CNN module has input 10 seconds long IR-UWB heartbeat signals, there is 360 sampled points.ECG signal is in input CNN mould
It is 200Hz by frequency 512Hz resampling before block, every 1200 sampled points are denoted as one section.Normalization factor and it is both configured to 1.
Signal acquisition working platform in order, is damaged coefficient and is both configured to 0.Experiment porch for training CNN is deep learning
Frame caffe.For specifically used network parameter as shown in Fig. 6, Fig. 7 and Fig. 8, Fig. 6 is provided in an embodiment of the present invention in experiment
The specific example structure chart of first CNN, Fig. 7 are the specific example structure chart and Fig. 8 of the 2nd CNN provided in an embodiment of the present invention
For the specific example structure chart of the 3rd CNN provided in an embodiment of the present invention.
It can be seen from the figure that three convolution kernels that the first CNN of the feature for extracting IR-UWB heartbeat signal is used
Size is respectively 1 × 61,1 × 31,1 × 21, and pond layer is all 1 × 2 to carry out average pond.For extracting ECG signal data
2nd CNN of feature has used three convolution kernels, and size is respectively 1 × 201,1 × 141,1 × 141, each pond layer is adopted
With 1 × 2 average pond method.20 1 × 20 characteristic patterns are finally both exported, are integrated into 2 × 20 feature, input the
Three CNN.The convolution kernel size that 3rd CNN is used be 2 × 5, pond layer be also 1 × 2 carry out be averaged pond.Two layer multi-layer senses
Know that the hidden layer of machine there are 50 neurons, output layer has 1 neuron.
In order to verify the accuracy for two classification (normal heartbeat, abnormal heartbeats) of this example setting, LCNN is used
Carry out performance comparison.Table 1 shows the result in different exception records and normal recordings ratio.This example uses following 3
A parameter measures classification performance.
Accuracy (Acc): the ratio of the ECG record and radar record correctly classified
Specific (Sp): the ratio that normal ECG record is correctly classified with normal radar record
Sensitivity (Se): the ratio that abnormal ECG record is correctly classified with abnormal radar record
The comparative experiments of 1 LCNN of table and cascade CNN
Although table 1 shows that the specificity of LCNN is more better than cascade CNN, LCNN sensitivity is 0, cannot identify exception
The shortcomings that ECG record, this also demonstrates LCNN, can not keep in the case where being exactly given training dataset in new data set
Stable performance.The sensitivity of cascade CNN is performed well.LCNN and cascade CNN accuracy be all with exception record with just
The often reduction of the ratio between record and improve, but the accuracy for cascading CNN is always higher, and fluctuation range is smaller.As it can be seen that even if number
Disequilibrium is still had according to collection, cascade CNN still is able to show high classification accuracy.
In second experiment, this example compares LCNN and cascades classification performance of the CNN under motion state.In data
In collection process, subject person is allowed to speak or weak vibrations body.As a result as shown in the table.
2 LCNN of table is compared with cascade CNN is in the classification performance under motion state
As it can be seen that cascade CNN the accuracy fluctuation within a narrow range near peak always, and with the static properties in table 1
Quite, it was demonstrated that its performance under light exercise state is stable.On the contrary, the accuracy of LCNN but maintain always it is lower
Level, and under having by a relatively large margin compared with the static properties in table 1 shows that LCNN is combined handling static and motion state
Under data when, performance be decline.The result again demonstrates, and cascade CNN uses IR-UWB heartbeat signal and ECG signal number
Satisfactory according to the arrhythmia classification effect of integrated analysis, especially under certain motion states, classification accuracy is still
It can achieve good level.
To sum up, the embodiment of the present invention, which is extracted and integrated by using cascade CNN, comes from ECG signal data and the IR-UWB heart
The feature of signal is jumped, more fully to be analyzed.Also just because of this integrated analysis, so that the embodiment of the present invention provides
System stable performance can be shown in the state of having light exercise.Pass through comparative experiments, it was demonstrated that implementation of the present invention
Example can achieve higher level in the accuracy of monitoring, even also having good stability table in light exercise
It is existing.
It is above to lift preferred embodiment, the object, technical solutions and advantages of the present invention are had been further described, institute
It should be understood that the foregoing is merely illustrative of the preferred embodiments of the present invention, it is not intended to limit the invention, it is all of the invention
Spirit and principle within, made any modifications, equivalent replacements, and improvements etc., should be included in protection scope of the present invention it
It is interior.
Claims (7)
1. a kind of removable electrocardiogram ECG monitors system characterized by comprising acquisition module and cascade CNN module, wherein
Acquisition module, for monitoring impulse radio ultra-wideband IR-UWB radar signal, synchronizing and being obtained after integrating processing
IR-UWB heartbeat signal；ECG signal is monitored, obtains ECG signal data after synchronizing processing and unbalanced data processing；
CNN module is cascaded, for carrying out to from the received IR-UWB heartbeat signal of acquisition module and treated ECG signal data
After feature extraction, integrated and diagnostic classification, monitoring result is obtained；
The cascade CNN module includes two layers:
First layer includes that two different CNN, the first CNN are extracted special for carrying out feature extraction to IR-UWB heartbeat signal
Sign；And the 2nd CNN, for treated, ECG signal data to carry out feature extraction；
The second layer is the 3rd CNN, and the feature for extracting to the first CNN and the 2nd CNN carries out classification mark, and label has normally
Heartbeat and two kinds of abnormal heartbeats；
Between first layer and the second layer include feature integrate submodule, two for extracting the first CNN and the 2nd CNN
Kind feature carries out after integrating, and is sent to the 3rd CNN；
The feature integrates submodule and uses weighting integrated, is expressed as follows:
Y=[a (1- θ1)x1, b (1- θ2)x2]
Wherein, y represents integrated 2D signal, x1Represent IR-UWB heartbeat signal, x2ECG signal is represented, a represents the IR-UWB heart
The normalization factor of signal is jumped, b represents the normalization factor of ECG signal, θ1∈ [0,1] represents the impaired of IR-UWB heartbeat signal
Coefficient, θ2∈ [0,1] represents the impaired coefficient of ECG signal.
2. monitoring system as described in claim 1, which is characterized in that the acquisition module further includes ECG signal acquisition submodule
Block, for monitoring ECG signal；The acquisition module further includes IR-UWB radar signal acquisition submodule, for monitoring IR-UWB
Radar signal.
3. monitoring system as described in claim 1, which is characterized in that the acquisition module is also used to using receiving time school
Quasi- mode synchronizes processing.
4. monitoring system as described in claim 1, which is characterized in that the acquisition module is in integrated processing IR-UWB radar letter
Number when, further include direct current component, band-pass filter unit, principal component analysis PCA removal noise signal unit, extract main letter
Number unit and integrated empirical mode decomposition unit, wherein
Direct current component is removed, for removing the DC component in IR-UWB radar signal；
Band-pass filter unit, for IR-UWB radar signal to be carried out bandpass filtering；
PCA removes noise signal unit, for removing the clutter in IR-UWB radar signal；
Main signal unit is extracted, for extracting the IR-UWB heartbeat signal in IR-UWB radar signal；
Integrated empirical mode decomposition unit is removed for that will extract obtained IR-UWB heartbeat signal rule of thumb Mode Decomposition
Breath signal therein finally obtains the IR-UWB heartbeat signal of needs.
5. monitoring system as described in claim 1, which is characterized in that the acquisition module further includes unbalanced data processing
Module increases ECG after carrying out unbalanced data processing using over-sampling mode and addition white noise mode to ECG signal
Abnormal heartbeats number in signal, obtains ECG signal data.
6. monitoring system as described in claim 1, which is characterized in that the first CNN or the 2nd CNN includes: 3 convolution
Layer, then 1 pond layer after each convolutional layer, 3 convolutional layers sequential connections, input data be all it is one-dimensional, convolutional layer
Convolution kernel be it is one-dimensional, in the forward propagation process, per one-dimensional convolutional layer calculation formula:
Wherein,It is l layers of kth characteristic pattern,It is the biasing of kth characteristic pattern in l layers,It is l-1 layers
I-th characteristic pattern,It is the convolution kernel mapped from l-1 layers of all features to l layers of k-th of characteristic pattern, N is
Total characteristic number in l-1 layers, symbol conv indicate that Vector convolution, f () indicate excitation function.
7. monitoring system as described in claim 1, which is characterized in that the 3rd CNN, including input layer, convolutional layer, Chi Hua
Layer, hidden layer and output layer, wherein
The feature of extraction is exported and is integrated by the first CNN and the 2nd CNN, and the two-dimensional matrix for forming the 3rd CNN is input to input layer,
After the convolutional layer for possessing two-dimensional convolution core by one, a pond layer is connect, then enters a hidden layer through full connection, most
After be that output layer provides monitoring result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610973366.6A CN106344005B (en) | 2016-10-28 | 2016-10-28 | A kind of removable electrocardiogram monitoring system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610973366.6A CN106344005B (en) | 2016-10-28 | 2016-10-28 | A kind of removable electrocardiogram monitoring system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106344005A CN106344005A (en) | 2017-01-25 |
| CN106344005B true CN106344005B (en) | 2019-04-05 |
Family
ID=57864972
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610973366.6A Active CN106344005B (en) | 2016-10-28 | 2016-10-28 | A kind of removable electrocardiogram monitoring system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106344005B (en) |
Families Citing this family (30)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108618773A (en) * | 2017-03-15 | 2018-10-09 | 深圳市理邦精密仪器股份有限公司 | A kind of Denoising of ECG Signal, device and a kind of ecg signal acquiring equipment |
| CN107358606B (en) * | 2017-05-04 | 2018-07-27 | 深圳硅基仿生科技有限公司 | The artificial neural network device and system and device of diabetic retinopathy for identification |
| CN107423571B (en) * | 2017-05-04 | 2018-07-06 | 深圳硅基仿生科技有限公司 | Diabetic retinopathy identifying system based on eye fundus image |
| CN107184198A (en) * | 2017-06-01 | 2017-09-22 | 广州城市职业学院 | A kind of electrocardiosignal classifying identification method |
| CN107529645B (en) * | 2017-06-29 | 2019-09-10 | 重庆邮电大学 | A kind of heart sound intelligent diagnosis system and method based on deep learning |
| CN107510452B (en) * | 2017-09-30 | 2019-10-08 | 扬美慧普（北京）科技有限公司 | A kind of ECG detecting method based on multiple dimensioned deep learning neural network |
| US10905328B2 (en) * | 2017-11-29 | 2021-02-02 | Verily Life Sciences Llc | Continuous detection and monitoring of heart arrhythmia using both wearable sensors and cloud-resident analyses |
| CN108030488A (en) * | 2017-11-30 | 2018-05-15 | 北京医拍智能科技有限公司 | The detecting system of arrhythmia cordis based on convolutional neural networks |
| CN107811649B (en) * | 2017-12-13 | 2020-12-22 | 四川大学 | Heart sound multi-classification method based on deep convolutional neural network |
| CN108256452A (en) * | 2018-01-06 | 2018-07-06 | 天津大学 | A kind of method of the ECG signal classification of feature based fusion |
| CN108577832B (en) * | 2018-02-11 | 2021-04-20 | 东南大学 | Computer readable storage medium for atrial fibrillation signal identification |
| CN108470158B (en) * | 2018-03-08 | 2020-05-12 | 华南理工大学 | Method for searching error minimum network computing structure for dynamic ECG data |
| CN108814580A (en) * | 2018-04-18 | 2018-11-16 | 深圳市康年颐华智能科技有限公司 | A kind of method and relevant device of non-contact type heart rate test |
| CN108814585B (en) * | 2018-05-03 | 2021-05-28 | 深圳竹信科技有限公司 | Electrocardiosignal processing method and device and computer readable storage medium |
| CN108919218A (en) * | 2018-06-07 | 2018-11-30 | 北京邮电大学 | A kind of contactless number of people in car and the method and device of position judgement |
| CN109044339A (en) * | 2018-08-13 | 2018-12-21 | 重庆工商大学 | A kind of ECG signal diagnostic method of limit convolutional neural networks |
| WO2020061987A1 (en) * | 2018-09-28 | 2020-04-02 | 深圳大学 | Multi-lead combined electrocardiogram classification method |
| CN109840511A (en) * | 2018-12-29 | 2019-06-04 | 深兰科技（上海）有限公司 | A kind of Modulation recognition method, terminal and storage medium |
| CN109829408B (en) * | 2019-01-23 | 2022-05-17 | 中国科学技术大学 | Intelligent lightning recognition device based on convolutional neural network |
| CN109875579A (en) * | 2019-02-28 | 2019-06-14 | 京东方科技集团股份有限公司 | Emotional health management system and emotional health management method |
| CN109965858B (en) * | 2019-03-28 | 2020-10-30 | 北京邮电大学 | Ultra-wideband radar-based human body vital sign detection method and device |
| CN110070013B (en) * | 2019-04-11 | 2021-06-18 | 杭州电子科技大学 | Intelligent evaluation method for noise pollution degree and category of ECG signal |
| CN110236520B (en) * | 2019-05-20 | 2020-09-29 | 上海数创医疗科技有限公司 | Electrocardiogram type recognition device based on double convolution neural network |
| CN111307881B (en) * | 2020-03-10 | 2022-03-08 | 南京工业大学 | Gas sensor baseline drift compensation method for VOC detection |
| CN111366184B (en) * | 2020-04-17 | 2022-03-01 | 中铁隧道局集团有限公司 | Shield tunneling machine multi-sensor performance online monitoring method |
| US11774553B2 (en) * | 2020-06-18 | 2023-10-03 | Infineon Technologies Ag | Parametric CNN for radar processing |
| CN111789574A (en) * | 2020-07-02 | 2020-10-20 | 山东省人工智能研究院 | ECG signal quality evaluation method |
| CN112130118B (en) * | 2020-08-19 | 2023-11-17 | 复旦大学无锡研究院 | Ultra-wideband radar signal processing system and method based on SNN |
| CN112232256B (en) * | 2020-10-26 | 2024-02-02 | 南京读动信息科技有限公司 | Contactless sports and body measurement data acquisition system |
| CN112971771B (en) * | 2021-02-23 | 2022-12-06 | 浙江大学计算机创新技术研究院 | Method for reconstructing electrocardiogram based on millimeter waves |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104970789A (en) * | 2014-04-04 | 2015-10-14 | 中国科学院苏州纳米技术与纳米仿生研究所 | Electrocardiogram classification method and system |
| CN105105743A (en) * | 2015-08-21 | 2015-12-02 | 山东省计算中心（国家超级计算济南中心） | Intelligent electrocardiogram diagnosis method based on deep neural network |
| CN105748063A (en) * | 2016-04-25 | 2016-07-13 | 山东大学齐鲁医院 | Intelligent arrhythmia diagnosis method based on multiple-lead and convolutional neural network |
| CN105943021A (en) * | 2016-05-13 | 2016-09-21 | 赵伟 | Wearable heart rhythm monitoring device and heart rhythm monitoring system |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| AU2003222001A1 (en) * | 2002-03-18 | 2003-10-08 | Sonomedica, Llc | Method and system for generating a likelihood of cardiovascular disease from analyzing cardiovascular sound signals. |
| GB0710963D0 (en) * | 2007-06-08 | 2007-07-18 | Psi Heartsignals Global Ltd | Methods of measurement of drug induced changes in cardiac ion channel function and associated apparatus |
-
2016
- 2016-10-28 CN CN201610973366.6A patent/CN106344005B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104970789A (en) * | 2014-04-04 | 2015-10-14 | 中国科学院苏州纳米技术与纳米仿生研究所 | Electrocardiogram classification method and system |
| CN105105743A (en) * | 2015-08-21 | 2015-12-02 | 山东省计算中心（国家超级计算济南中心） | Intelligent electrocardiogram diagnosis method based on deep neural network |
| CN105748063A (en) * | 2016-04-25 | 2016-07-13 | 山东大学齐鲁医院 | Intelligent arrhythmia diagnosis method based on multiple-lead and convolutional neural network |
| CN105943021A (en) * | 2016-05-13 | 2016-09-21 | 赵伟 | Wearable heart rhythm monitoring device and heart rhythm monitoring system |
Non-Patent Citations (1)
| Title |
|---|
| 《心电信号波形检测与心律失常分类研究》;王利琴;《中国博士学位论文全文数据库》;20160115(第01期) |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106344005A (en) | 2017-01-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106344005B (en) | A kind of removable electrocardiogram monitoring system | |
| Yin et al. | ECG monitoring system integrated with IR-UWB radar based on CNN | |
| US11311201B2 (en) | Feature selection for cardiac arrhythmia classification and screening | |
| US9974488B2 (en) | Early detection of hemodynamic decompensation using taut-string transformation | |
| CN102499664B (en) | Video-image-based method and system for detecting non-contact vital sign | |
| CN103034837B (en) | Characteristic parameter is associated with pulse condition key element | |
| CN111466876B (en) | Alzheimer's disease auxiliary diagnosis system based on fNIRS and graph neural network | |
| US20150230756A1 (en) | Determining physiological characteristics from sensor signals including motion artifacts | |
| CN109907752A (en) | A kind of cardiac diagnosis and monitoring method and system of the interference of removal motion artifacts and ecg characteristics detection | |
| CN105125206B (en) | A kind of intelligent cardiac monitoring method and device | |
| Zhao et al. | An IoT-based wearable system using accelerometers and machine learning for fetal movement monitoring | |
| CN105249925B (en) | A kind of traditional Chinese medical pulse manifestation collecting device and noise reduction system and noise-reduction method | |
| CN113095302B (en) | Depth model for arrhythmia classification, method and device using same | |
| CN111887858B (en) | Ballistocardiogram signal heart rate estimation method based on cross-modal mapping | |
| CN110448281A (en) | A kind of wearable work fatigue detection system based on multisensor | |
| CN104571504B (en) | A kind of online brain-machine interface method based on Imaginary Movement | |
| Wang et al. | A novel multi-scale dilated 3D CNN for epileptic seizure prediction | |
| CN112806977B (en) | Physiological parameter measuring method based on multi-scale fusion network | |
| Orguc et al. | EMG-based real time facial gesture recognition for stress monitoring | |
| CN109674464A (en) | A kind of multi-lead electrocardiosignal compound characteristics extracting method and corresponding monitoring system | |
| Kumar et al. | STSR: spectro-temporal super-resolution analysis of a reference signal less photoplethysmogram for heart rate estimation during physical activity | |
| CN109770920A (en) | Intense strain method of discrimination and its system based on wearable ECG signal | |
| CN113812933A (en) | Acute myocardial infarction real-time early warning system based on wearable equipment | |
| Pathan et al. | A machine learning based human activity recognition during physical exercise using wavelet packet transform of ppg and inertial sensors data | |
| CN106419884A (en) | Heart rate calculating method and system based on wavelet analysis |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |