CN113796873B - Wearable dynamic electrocardiosignal classification method and system - Google Patents
Wearable dynamic electrocardiosignal classification method and system Download PDFInfo
- Publication number
- CN113796873B CN113796873B CN202111247597.6A CN202111247597A CN113796873B CN 113796873 B CN113796873 B CN 113796873B CN 202111247597 A CN202111247597 A CN 202111247597A CN 113796873 B CN113796873 B CN 113796873B
- Authority
- CN
- China
- Prior art keywords
- scattering
- wavelet
- order
- signals
- electrocardiosignals
- 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
- 238000000034 method Methods 0.000 title claims abstract description 39
- 230000006870 function Effects 0.000 claims abstract description 56
- 239000011159 matrix material Substances 0.000 claims abstract description 45
- 238000013145 classification model Methods 0.000 claims abstract description 24
- 238000012545 processing Methods 0.000 claims abstract description 23
- 238000012549 training Methods 0.000 claims description 42
- 239000006185 dispersion Substances 0.000 claims description 21
- 238000004590 computer program Methods 0.000 claims description 18
- 238000000605 extraction Methods 0.000 claims description 18
- 230000015654 memory Effects 0.000 claims description 18
- 238000013441 quality evaluation Methods 0.000 claims description 18
- 201000010099 disease Diseases 0.000 claims description 14
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 claims description 14
- 238000003745 diagnosis Methods 0.000 claims description 12
- 238000002372 labelling Methods 0.000 claims description 12
- 238000001514 detection method Methods 0.000 claims description 10
- 238000005070 sampling Methods 0.000 claims description 10
- 238000001303 quality assessment method Methods 0.000 claims description 9
- 238000001228 spectrum Methods 0.000 claims description 9
- 238000003860 storage Methods 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 abstract description 14
- 238000010801 machine learning Methods 0.000 abstract description 5
- 238000004458 analytical method Methods 0.000 abstract description 4
- 238000010586 diagram Methods 0.000 description 7
- 230000000694 effects Effects 0.000 description 6
- 238000012360 testing method Methods 0.000 description 5
- 230000009466 transformation Effects 0.000 description 5
- 238000013135 deep learning Methods 0.000 description 4
- 238000011156 evaluation Methods 0.000 description 4
- 230000035945 sensitivity Effects 0.000 description 4
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 description 3
- 208000024172 Cardiovascular disease Diseases 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 3
- 230000006403 short-term memory Effects 0.000 description 3
- 238000013519 translation Methods 0.000 description 3
- 230000002457 bidirectional effect Effects 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 230000007787 long-term memory Effects 0.000 description 2
- 230000007774 longterm Effects 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000002526 effect on cardiovascular system Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000877 morphologic effect Effects 0.000 description 1
- 230000003183 myoelectrical effect Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000011425 standardization method Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
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
- A61B5/68—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
- A61B5/6802—Sensor mounted on worn items
-
- 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
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
- A61B5/7267—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems involving training the classification device
-
- 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)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Animal Behavior & Ethology (AREA)
- Veterinary Medicine (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Molecular Biology (AREA)
- Surgery (AREA)
- Artificial Intelligence (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Pathology (AREA)
- Signal Processing (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physiology (AREA)
- Psychiatry (AREA)
- Cardiology (AREA)
- Evolutionary Computation (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention provides a wearable dynamic electrocardiosignal classification method and system, which belong to the technical field of electrocardiosignal analysis and acquire an original electrocardiosignal to be classified; extracting wavelet scattering coefficients of electrocardiosignals to be classified by utilizing a pre-constructed wavelet scattering network, calculating a scattering feature matrix by utilizing the wavelet scattering coefficients, and converting the scattering feature matrix into a scattering feature map; and processing the scattering feature map by using a pre-trained classification model to obtain a classification result of the electrocardiosignal. The wearable dynamic signals are divided into three types, and a signal classification database is constructed; the judgment criteria of lead falling and pure noise are provided, and the calculation cost is reduced; the method comprises the steps of constructing a wavelet scattering network by using a Gabor wavelet function and a scale operator, extracting wavelet scattering coefficients from three types of signals by using the wavelet scattering network, constructing a scattering characteristic matrix, extracting scattering matrix characteristics by using a machine learning method LSTM and classifying, thereby realizing automatic and accurate classification of wearable dynamic electrocardiosignals.
Description
Technical Field
The invention relates to the technical field of electrocardiosignal analysis, in particular to a wearable dynamic electrocardiosignal classification method and system.
Background
For early prevention and continuous monitoring of cardiovascular diseases, in-and-out noise and motion artifact in different motion states are varied in a dynamic long-time monitoring environment, signal noise is serious, the type and degree of noise are unpredictable, the quality of wearable signals is complex and changeable and difficult to control, and the wearable signals are easily interfered by body movement, environmental noise, electromagnetism and the like, so that the signal quality is reduced, and false alarm is generated. Redundant information and high false alarm rate in massive data not only lead to medical resource waste, but also lead to the fact that doctors are in a poor condition and even misdiagnosis is caused. Therefore, the dynamic physiological data must be subjected to quality evaluation firstly, and the dynamic physiological data are divided into signals which can not be used and can be used for clinic, so that the purposes of removing coarse and remaining essence and removing false and passbook are achieved, and the efficiency and the diagnosis quality are improved.
Currently, the electrocardiosignal quality evaluation is mainly based on the extraction of quality evaluation indexes, and the quality evaluation index extraction mainly relates to time domain, frequency domain, nonlinear domain, spatial morphological information and the like. The frequency domain measure set forth by the Murray subject group at NewCastle university, langley et al uses 6 waveform features of straight line, baseline elevation, baseline drift, too low amplitude, too high amplitude, kurtosis, etc. for quality assessment, zaunseder et al in combination with the power spectrum set forth 35 frequency domain indicators. These methods often have difficulty in reasonably setting various threshold parameters in practical applications, resulting in weak generalization capability when applied to different scenes and different leads.
Aiming at the processing of electrocardiosignals, the wavelet transformation can effectively analyze nonstationary electrocardiosignals by utilizing the time-frequency domain accurate positioning characteristic, but the stability of the scale transformation cannot be maintained.
Disclosure of Invention
The invention aims to provide a wearable dynamic electrocardiosignal classification method and system based on a wavelet dispersion network, which are used for solving at least one technical problem in the background technology.
In order to achieve the above purpose, the present invention adopts the following technical scheme:
In one aspect, the invention provides a wearable dynamic electrocardiosignal classification method, which comprises the following steps:
acquiring an original electrocardiosignal to be classified;
Extracting wavelet scattering coefficients of electrocardiosignals to be classified by utilizing a pre-constructed wavelet scattering network, calculating a scattering feature matrix by utilizing the wavelet scattering coefficients, and converting the scattering feature matrix into a scattering feature map;
processing the scattering feature map by using a pre-trained classification model to obtain a classification result of the electrocardiosignal; the pre-trained classification model is trained by a training set, and the training set comprises a plurality of electrocardiosignals and labels for labeling the classes of the electrocardiosignals.
Preferably, a three-layer wavelet scattering network is constructed based on the scale function and the wavelet function, and 0-order scattering coefficient, 1-order scattering coefficient and 2-order scattering coefficient are generated.
Preferably, the electrocardiosignals to be classified are convolved with a scale function to obtain a 0-order wavelet dispersion coefficient; the electrocardiosignals to be classified are convolved with a first-order wavelet function, and a first-order scattering propagation operator is generated through nonlinear modulo operation; convolving the first-order scattering propagation operator with the scale function to obtain a first-order wavelet scattering coefficient; the first-order scattering propagation operator is convolved with a second-order wavelet function, and a second-order scattering propagation operator is generated through nonlinear modulo operation; the second-order scattering propagation operator is convolved with the scale function to obtain a second-order wavelet scattering coefficient.
Preferably, the maximum scale factor by which the scale function remains translationally invariant, i.e. the instant support, is determined based on the length of the signal and the sampling frequency.
Preferably, the basic network for training the classification model is a long-term and short-term memory neural network.
Preferably, the training set is from a quality evaluation database, wherein the quality evaluation database comprises a plurality of electrocardiosignals which are manually marked, and the category of the electrocardiosignals comprises clean signals which can be used for disease detection and diagnosis, light pollution signals which can be used for heart rate extraction and noise pollution serious signals which need to be removed.
Preferably, all data in the quality assessment database is first preprocessed, including: data normalization processing, lead drop judgment and pure noise judgment.
In a second aspect, the present invention provides a wearable dynamic electrocardiograph signal classification system, comprising:
the acquisition module is used for acquiring the original electrocardiosignals to be classified;
the extraction module is used for extracting wavelet scattering coefficients of the electrocardiosignals to be classified by utilizing a pre-constructed wavelet scattering network, calculating a scattering feature matrix by utilizing the wavelet scattering coefficients and converting the scattering feature matrix into a scattering feature map;
The classification module is used for processing the scattering feature map by utilizing a pre-trained classification model to obtain a classification result of the electrocardiosignal; the pre-trained classification model is trained by a training set, and the training set comprises a plurality of electrocardiosignals and labels for labeling the classes of the electrocardiosignals.
In a third aspect, the present invention provides a non-transitory computer readable storage medium for storing computer instructions which, when executed by a processor, implement a wearable dynamic electrocardiograph signal classification method as described above.
In a fourth aspect, the present invention provides an electronic device comprising: a processor, a memory, and a computer program; wherein the processor is connected to the memory, and the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory, so that the electronic device executes the instructions for implementing the wearable dynamic electrocardiosignal classification method as described above.
The invention has the beneficial effects that:
Wearable dynamic signals are divided into three categories: the device can be used for detecting and diagnosing diseases, and can be used for detecting and diagnosing diseases, namely relatively clean signals, light pollution signals only used for heart rate extraction and signals which are seriously polluted by noise and need to be removed; constructing a wavelet scattering network by using a Gabor wavelet function and a scale operator, extracting wavelet scattering coefficients by using the wavelet scattering network, and constructing a scattering feature matrix by using the scattering coefficients; the judgment criteria of lead falling and pure noise are provided, so that the calculation cost is greatly reduced; three types of signals are extracted by wavelet scattering to obtain a scattering feature matrix, and the features of the scattering matrix are extracted by a machine learning method LSTM and classified, so that automatic and accurate classification of wearable dynamic electrocardiosignals is realized.
Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flowchart of a method for classifying wearable dynamic electrocardiographic signals based on a wavelet dispersion network according to an embodiment of the present invention.
Fig. 2 is a schematic diagram of an original electrocardiograph signal and an electrocardiograph signal after noise is added according to an embodiment of the present invention.
Detailed Description
Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to like or similar elements throughout or elements having like or similar functionality. The embodiments described below by way of the drawings are exemplary only and should not be construed as limiting the invention.
It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.
As used herein, the singular forms "a", "an", "the" and "the" are intended to include the plural forms as well, unless expressly stated otherwise, as understood by those skilled in the art. It will be further understood that the terms "comprises" and/or "comprising," when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, and/or groups thereof.
In the description of the present specification, a description referring to terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present invention. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, the different embodiments or examples described in this specification and the features of the different embodiments or examples may be combined and combined by those skilled in the art without contradiction.
In order that the invention may be readily understood, a further description of the invention will be rendered by reference to specific embodiments that are illustrated in the appended drawings and are not to be construed as limiting embodiments of the invention.
It will be appreciated by those skilled in the art that the drawings are merely schematic representations of examples and that the elements of the drawings are not necessarily required to practice the invention.
Example 1
The wavelet dispersion network with the depth convolution network is constructed by utilizing nonlinear modulo operation and scaling operator cascading wavelet transformation, the wavelet dispersion transformation has very high resolution in a time-frequency domain, has computational translation invariance, can keep higher stability for deformation, and can keep high-frequency information. In addition, wavelet scattering networks have the characteristics of deep learning architecture when extracting data features: multi-scale shrinkage, linearization of hierarchical symmetry, and coefficient characterization. The wavelet dispersion network maintains the advantages of the traditional method, integrates the characteristics of the deep learning network, and has higher characteristics in the fields of pattern recognition, audio analysis, signal processing and the like. Thus, the wavelet dispersion network has great advantages for the electrocardiographic quality assessment work mixed with various complex noises.
Therefore, the present embodiment 1 provides a wearable dynamic electrocardiograph signal classification system and method based on a wavelet dispersion network.
First, the wearable dynamic electrocardiograph signal classification system provided in this embodiment 1 includes:
the acquisition module is used for acquiring the original electrocardiosignals to be classified;
the extraction module is used for extracting wavelet scattering coefficients of the electrocardiosignals to be classified by utilizing a pre-constructed wavelet scattering network, calculating a scattering feature matrix by utilizing the wavelet scattering coefficients and converting the scattering feature matrix into a scattering feature map;
The classification module is used for processing the scattering feature map by utilizing a pre-trained classification model to obtain a classification result of the electrocardiosignal; the pre-trained classification model is trained by a training set, and the training set comprises a plurality of electrocardiosignals and labels for labeling the classes of the electrocardiosignals.
By using the wearable dynamic electrocardiosignal classification system, the wearable dynamic electrocardiosignal classification method is realized, and comprises the following steps:
acquiring an original electrocardiosignal to be classified;
Extracting wavelet scattering coefficients of electrocardiosignals to be classified by utilizing a pre-constructed wavelet scattering network, calculating a scattering feature matrix by utilizing the wavelet scattering coefficients, and converting the scattering feature matrix into a scattering feature map;
processing the scattering feature map by using a pre-trained classification model to obtain a classification result of the electrocardiosignal; the pre-trained classification model is trained by a training set, and the training set comprises a plurality of electrocardiosignals and labels for labeling the classes of the electrocardiosignals.
In this embodiment 1, a three-layer wavelet scattering network was constructed based on a scale function and a wavelet function, generating 0-order scattering coefficient, 1-order scattering coefficient, and 2-order scattering coefficient.
Specifically, an electrocardiosignal to be classified is convolved with a scale function to obtain a 0-order wavelet dispersion coefficient; the electrocardiosignals to be classified are convolved with a first-order wavelet function, and a first-order scattering propagation operator is generated through nonlinear modulo operation; convolving the first-order scattering propagation operator with the scale function to obtain a first-order wavelet scattering coefficient; the first-order scattering propagation operator is convolved with a second-order wavelet function, and a second-order scattering propagation operator is generated through nonlinear modulo operation; the second-order scattering propagation operator is convolved with the scale function to obtain a second-order wavelet scattering coefficient.
When constructing a wavelet dispersion network, determining the maximum scale factor of which the scale function keeps the translational invariance, namely the real-time support degree based on the length and the sampling frequency of a signal.
In this embodiment 1, the basic network used for training the classification model is a long-short term memory neural network.
The training set is from a quality evaluation database, wherein the quality evaluation database comprises a plurality of electrocardiosignals which are marked manually, and the types of the electrocardiosignals comprise clean signals which can be used for disease detection and diagnosis, light pollution signals which can be used for heart rate extraction and noise pollution serious signals which need to be removed. The long-time dynamic wearable dynamic electrocardiosignals are segmented by utilizing a10 s sliding window, and the problem of data imbalance is solved by adding noise.
All data in the quality assessment database are first preprocessed, including: data normalization processing, lead drop judgment and pure noise judgment.
The segmented 10s signal is firstly subjected to judgment of whether the lead falls off, if the lead falls off signal, the lead is directly removed, and the program is finished, so that the calculation cost can be saved; and then judging the pure Gaussian noise, if the pure Gaussian noise signal is the pure Gaussian noise signal, directly eliminating the pure Gaussian noise signal, and ending the program, so that the calculation cost can be saved.
Example 2
Because of body movement, environmental noise, electromagnetic interference and the like, the quality of the wearable dynamic electrocardiosignal is complex and changeable and is difficult to control, great interference is brought to doctor diagnosis, and how to evaluate the quality of the wearable electrocardiosignal, remove coarse and remain essence and remove false passbook is a key problem to be solved urgently for improving early cardiovascular detection and diagnosis efficiency. In embodiment 2 of the invention, a classification method of wearable dynamic electrocardiosignals is provided, the advantages of a wavelet scattering network are fully utilized, scattering characteristics of the wearable dynamic electrocardiosignals are extracted, a scattering characteristic matrix is constructed, signals are classified by using a Long Short-Term Memory (LSTM) network, signals with serious noise pollution are removed based on classification results, reserved signals are divided into mild pollution signals only used for heart rate extraction and clean signals used for disease classification, and more powerful data is provided for real-time accurate detection of cardiovascular diseases.
In the classification method of wearable dynamic electrocardiograph signals in embodiment 2, a database of clean signals, slightly polluted signals and severely polluted signals needs to be established, and a proper wavelet scattering network needs to be established, so that scattering feature matrix characteristics of various data in the database are accurately obtained, and meanwhile, evaluation indexes of quality classification effects of three types of data need to be determined, and classification effects of a model are evaluated.
In this embodiment 2, a wavelet scattering network including 41 1-order wavelet functions and 7 2-order wavelet functions is built by using Gabor wavelet functions and a scaling operator, 0-order, 1-order and 2-order wavelet scattering coefficients are extracted by using the wavelet scattering network, an 81×20-dimensional scattering feature matrix is built by using the scattering coefficients, and finally features in the scattering feature map are extracted and classified by using a deep learning LSTM method, so that wearable dynamic signals are classified into three types, namely, noise pollution serious signals which need to be removed, mild pollution signals which can be used for heart rate extraction, and relatively clean signals which can be used for disease detection and diagnosis.
The wavelet scattering has the characteristics of translational invariance, characteristic stability, information richness and the like, and the characteristics enable the wavelet scattering to have higher discernment on the deformation of the wearable dynamic electrocardiosignal. In this embodiment 2, constructing a three-layer wavelet dispersion network mainly includes the following steps:
And selecting a scale function phi I with a time support of I (Invariance scale) and a Morlet wavelet function phi to construct a three-layer wavelet scattering network, generating 0-order, 1-order and 2-order scattering coefficients, wherein the network can cover the whole frequency domain range of the signal. The network construction steps are as follows:
(1) X (t) represents an electrocardiosignal to be analyzed, and is convolved with a scale function phi I to obtain a 0-order wavelet dispersion coefficient S 0：S0X(t)＝X(t)*φI;
(2) Electrocardiosignal X (t) and first-order wavelet function Convolving and generating a first-order scattering propagation operator/>, through nonlinear modulo operation
(3) The first-order propagation operator is convolved with the scale function to obtain a first-order wavelet dispersion coefficient S 1:
(4) First order scatter propagation operator And second order wavelet function/>Convolving and generating a second-order scattering propagation operator/>, through nonlinear modulo operation
(5) The second-order scattering propagation operator is convolved with the scale function to obtain a second-order wavelet scattering coefficient S 2:
In order to achieve accurate classification of three types of quality data, in this embodiment 2, a database is constructed, and data in the database is preprocessed: the long-time dynamic wearable dynamic electrocardiosignals are segmented by utilizing a10 s sliding window, and the problem of data imbalance is solved by adding noise. The segmented 10s signal is firstly subjected to judgment of whether the lead falls off, if the lead falls off signal, the lead is directly removed, and the program is finished, so that the calculation cost can be saved; and then judging the pure Gaussian noise, if the pure Gaussian noise signal is the pure Gaussian noise signal, directly eliminating the pure Gaussian noise signal, and ending the program, so that the calculation cost can be saved.
And calculating a wavelet scattering feature matrix through the three-layer wavelet scattering network constructed by the two judged signal inputs. Firstly, determining the time support degree I of a scale function, namely, the maximum scale factor for keeping the translation invariance, determining a proper I value based on the length of a signal and the sampling frequency Fs, and proving that the I generally takes the value of 2-10 xFs through experiments. And calculating 0-order, 1-order and 2-order scattering coefficients, extracting the scattering coefficients of the three-layer wavelet network to construct a scattering feature matrix, and converting the scattering feature matrix into a scattering feature map.
The scattering feature matrix extracted by the training data wavelet scattering network is input into a long-short-term memory neural network LSTM, and a solver adopts an Adam algorithm to determine proper parameters such as gradient threshold, maximum round number, small batch number and the like. The dimension of the sequence input layer is 81 multiplied by 20, hidden nodes are designated through the bidirectional LSTM layer, the last classification value is output, then the sequence input layer enters the full-connection layer, the output category is pointed out, the softmax layer is arranged behind the output category, the probability of classification of each category is output, and finally the final classification result is output through the classification layer.
In the method for evaluating the wearable dynamic electrocardiographic quality based on the wavelet dispersion network in the embodiment 2, the wearable dynamic signals are divided into A, B, C types, wherein the A type is a relatively clean signal which can be used for disease detection and diagnosis, the B type is a slightly polluted signal which can only be used for heart rate extraction, and the C type is a signal which is seriously polluted by noise and needs to be removed. And constructing a wavelet scattering network of 0 order, 1 order and 2 order by using a Gabor wavelet function and a scaling operator, extracting 3 layers of wavelet scattering coefficients by using the wavelet scattering network, constructing a scattering feature matrix by using the scattering coefficients, and finally extracting and classifying the features of the scattering matrix by using a deep learning LSTM method.
The current research on wearable dynamic electrocardiographic quality assessment mostly classifies data quality grades into two types, three types of signals are classified in the embodiment 2, and a quality assessment database is constructed, wherein the database comprises 10s signals of data 31111, 11709 signals of clean signals, 7860 signals of light pollution and 11542 signals of serious pollution. The proposed judgment criteria of lead falling and pure noise greatly reduce the calculation cost. The method comprises the steps of utilizing a sharp tool for wavelet scattering signal analysis to extract scattering feature matrixes of A, B, C types of signals, and then utilizing a machine learning method LSTM to extract the features of the scattering matrixes and classify the scattering matrixes, so that the automatic classification of wearable dynamic electrocardiograph data is realized. 31111 data in the existing database are trained and tested, and the accuracy is about 95%.
Example 3
In this embodiment 3, a method for classifying dynamic electrocardiographic signals acquired by a wearable device based on a wavelet dispersion network is provided, and as shown in fig. 1, the main steps are divided into four steps. Firstly, constructing a wearable dynamic electrocardiographic quality evaluation database, dividing long-term dynamic electrocardiographic into signals with the length of 10s according to information marked by an expert, and determining a label of each signal; the second step is to preprocess the data, firstly, standardized processing is carried out, then, the signal of lead falling and the pure Gaussian noise signal are removed, and the calculation cost is reduced; thirdly, inputting the residual signals into a constructed wavelet scattering network to generate a scattering feature matrix; and fourthly, training the LSTM network by using 70% of data of the database as training data, and using 30% of data as test data to test the classification effect of the model.
In this embodiment 3, constructing the wearable dynamic electrocardiographic quality assessment database includes:
The length of each data set is about 24 hours, the data comes from 15 testers (men 6 and women 9), the age distribution is 21-83 years, the electrocardiosignals of the testers in the daily living environment are recorded by using wearable electrocardiosignal equipment, and the sampling frequency of the signals is 1000Hz. In order to reduce the calculation cost, in this embodiment, the down-sampling frequency of the signal is 250Hz, the long-term dynamic signal is divided into signals with the length of 10s according to the quality labeling information labeled by the expert, as shown in table 1, and the quality label of each signal is determined. Signal quality classes are divided into three classes: class a is a relatively clean set of 11709 signals that can be used for disease detection and diagnosis, class B is a set of 7860 lightly contaminated signals that can be used for heart rate extraction, and class C is a set of 2687 severely contaminated noise signals that need to be removed.
Because the noise data size is smaller, the B-type signals are randomly divided into 6 groups, six types of noise (limit offset 2 groups, myoelectric noise 2 groups and electrode movement noise 2 groups) are added, 1000 pieces of data are randomly selected from the A-type data, gaussian noise is added, and the noise added data and the original 2687 pieces of noise data form 11542 pieces of noise data together. Three types of data are 31111 pieces in total to form a signal quality evaluation database, as shown in fig. 2, N1-N6 are 6 signals from class B light pollution, M1-6 are C serious pollution signals which are obtained by adding 6 kinds of noise to the 6 signals, N7 is a clean signal from class a, and M7 is a serious pollution signal which is obtained by adding gaussian noise to the 6 signals.
TABLE 1 three types of signals for database
All data in the quality evaluation database are preprocessed first, and the data preprocessing stage is divided into three parts. A set of ECG signals s= { S 1,s2,...,sn-1,sn } is agreed to have a total of n sample numbers, the first order difference diff (S) = { S 2-s1,s3-s2,...,sn-sn-1 } of which signals.
Data standardization processing:
In order to eliminate the influence among different types of noise, firstly, data standardization processing is needed, and the original data is mapped between [0-1] in a linear transformation way by using a min-max standardization method, wherein the formula is as follows:
Judging lead falling:
And the standardized 10s signal is used for judging the lead falling, if the lead falling signal is directly removed, the program is finished, and the calculation cost can be saved. For a signal, if the number of samples with equal amplitude of two adjacent samples is greater than 60% of the total number of samples, the signal is considered to be a lead drop signal, and the calculation formula is as follows:
for sample s i:
Pure noise judgment:
And thirdly, judging the pure noise, if the pure noise signals are directly removed, ending the program, and saving the calculation cost. Based on the electrocardiosignal spectrum range of 0-40Hz, if the ratio of the power spectrum energy of the signal in the range of 0-40Hz to the total energy is less than 30%, the main component of the signal is not the electrocardiosignal, belongs to the noise signal and can be directly abandoned. The calculation formula is as follows:
in this embodiment 3, constructing a wavelet dispersion network to calculate a wavelet dispersion feature matrix includes:
And calculating a wavelet scattering feature matrix through the three-layer wavelet scattering network constructed by the two judged signal inputs. Firstly, determining the time support degree I of a scale function, namely the maximum scale factor for keeping the translation invariance, determining a proper I value based on the length of a signal and the sampling frequency Fs, and proving that the I generally takes the value of 2-10 xFs through experiments. And calculating 0-order, 1-order and 2-order scattering coefficients, and extracting the scattering coefficients of the three-layer wavelet network to construct a scattering feature matrix. The network construction steps are as follows:
(1) X (t) represents an electrocardiosignal to be analyzed, and is convolved with a scale function phi I to obtain a 0-order wavelet dispersion coefficient S 0：S0X(t)＝X(t)*φI;
(2) Electrocardiosignal X (t) and first-order wavelet function Convolving and generating a first-order scattering propagation operator/>, through nonlinear modulo operation
(3) The first-order propagation operator is convolved with the scale function to obtain a first-order wavelet dispersion coefficient S 1:
(4) First order scatter propagation operator And second order wavelet function/>Convolving and generating a second-order scattering propagation operator/>, through nonlinear modulo operation
(5) The second-order scattering propagation operator is convolved with the scale function to obtain a second-order wavelet scattering coefficient S 2:
In this embodiment 3, the creating of the classification model using the machine learning long-term memory network LSTM includes:
And inputting a scattering feature matrix extracted by a training data wavelet scattering network into the LSTM, and determining proper parameters such as gradient threshold, maximum round number, small batch number and the like by adopting an Adam algorithm by a solver. The dimension of the sequence input layer is 81 multiplied by 20, hidden nodes are designated through the bidirectional LSTM layer, the last classification value is output, then the sequence input layer enters the full-connection layer, the output category is pointed out, the softmax layer is arranged behind the output category, the probability of classification of each category is output, and finally the final classification result is output through the classification layer.
In this example 3, the total number of databases was 31111, 70% of the data was randomly selected as training data, and the remaining 30% was used as test data. The training process setting parameters are set as follows, adam is adopted as a solver, maxEpoch is 1000, minBasize is 1000, initialLength is 0.001, and the execution environment is a single GPU.
Table 2 gives the confusion matrix for the test results. The test training set totals 9333 sets of data, wherein the class a data is 3513 sets, the class B data is 2352 sets, and the class C data is 3495 sets.
In this example 3, the evaluation index used was sensitivity SENSITIVIETY (SE), accuracy (+p), three quality levels of accuracy and sensitivity balance combination index F 1 measure: f 1A,F1B,F1C, and an Accuracy (ACC).
Sensitivity Se: the proportion of the true predicted correct samples of a certain type to the total number of all samples of a certain type in the training set is exemplified by the samples of a type:
accuracy + P: the true predicted correct samples of a certain class are proportional to the total number of all predicted samples of a certain class in the training set. Taking class a samples as an example:
TABLE 2
Accuracy and sensitivity balance comprehensive index F 1 measure:
Wherein TN A、TNB、TNC is the number of groups accurately predicted as A, B, C class signals respectively, N A、NB、NC is the number of groups accurately predicted as A, B, C class signals in the training set, and T A、TB、TC is the number of groups accurately predicted as A, B, C class signals in the training set.
The accuracy represents the proportion of all types of samples which are truly predicted to be correct to the total number of all samples in the training set, and the calculation formula is as follows:
finally, the classification results of the three quality class signals are shown in table 3.
TABLE 3 Table 3
In summary, in the embodiment 3, the proposed wearable electrocardiograph signal classification method constructs a wearable dynamic electrocardiograph quality evaluation database including three quality classes; determining a judging standard of the lead falling and the pure noise signal; constructing a three-layer wavelet scattering network suitable for extracting a wearable dynamic electrocardiograph scattering matrix; determining reasonable scattering paths suitable for three quality grade electrocardiosignal characteristic extraction; determining evaluation indexes suitable for the classification effect of the electrocardiosignals with three quality grades; finally, the characteristics of the dynamic electrocardiograph scattering matrixes with three quality grades are extracted A, B, C by using a machine learning method LSTM, three classifications of the wearable dynamic electrocardiograph data quality are realized, and the classification accuracy reaches 95.44%. And the data with serious pollution are removed, clean signals suitable for disease diagnosis are screened, and diagnosis time is saved for doctors.
Example 4
Embodiment 4 of the present invention provides a non-transitory computer readable storage medium for storing computer instructions, which when executed by a processor, implement a wearable dynamic electrocardiograph signal classification method as described above, the method comprising:
acquiring an original electrocardiosignal to be classified;
Extracting wavelet scattering coefficients of electrocardiosignals to be classified by utilizing a pre-constructed wavelet scattering network, calculating a scattering feature matrix by utilizing the wavelet scattering coefficients, and converting the scattering feature matrix into a scattering feature map;
processing the scattering feature map by using a pre-trained classification model to obtain a classification result of the electrocardiosignal; the pre-trained classification model is trained by a training set, and the training set comprises a plurality of electrocardiosignals and labels for labeling the classes of the electrocardiosignals.
Example 5
Embodiment 5 of the present invention provides a computer program (product) comprising a computer program for implementing a wearable dynamic electrocardiographic signal classification method as described above when run on one or more processors, the method comprising:
acquiring an original electrocardiosignal to be classified;
Extracting wavelet scattering coefficients of electrocardiosignals to be classified by utilizing a pre-constructed wavelet scattering network, calculating a scattering feature matrix by utilizing the wavelet scattering coefficients, and converting the scattering feature matrix into a scattering feature map;
processing the scattering feature map by using a pre-trained classification model to obtain a classification result of the electrocardiosignal; the pre-trained classification model is trained by a training set, and the training set comprises a plurality of electrocardiosignals and labels for labeling the classes of the electrocardiosignals.
Example 6
Embodiment 6 of the present invention provides an electronic device, including: a processor, a memory, and a computer program; wherein the processor is connected to the memory, and the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory, so that the electronic device executes the instructions for implementing the wearable dynamic electrocardiograph signal classification method, and the method includes:
acquiring an original electrocardiosignal to be classified;
Extracting wavelet scattering coefficients of electrocardiosignals to be classified by utilizing a pre-constructed wavelet scattering network, calculating a scattering feature matrix by utilizing the wavelet scattering coefficients, and converting the scattering feature matrix into a scattering feature map;
processing the scattering feature map by using a pre-trained classification model to obtain a classification result of the electrocardiosignal; the pre-trained classification model is trained by a training set, and the training set comprises a plurality of electrocardiosignals and labels for labeling the classes of the electrocardiosignals.
In summary, the wearable dynamic electrocardiosignal classification method and system fully utilize the advantages of the wavelet scattering network, extract scattering characteristics of the wearable dynamic electrocardiosignal, construct a scattering characteristic matrix, classify signals by using a Long Short-Term Memory (LSTM) network, kick out signals with serious noise pollution based on classification results, and divide reserved signals into light pollution signals only used for heart rate extraction and clean signals used for disease classification, thereby providing more powerful data for real-time accurate detection of cardiovascular diseases. The method needs to establish a clean signal database, a light pollution signal database and a serious pollution signal database, and needs to establish a proper wavelet scattering network so as to accurately acquire the scattering feature matrix characteristics of various data in the database, and meanwhile, needs to determine the evaluation indexes of the quality classification effects of the three types of data and evaluate the classification effects of a model.
It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While the foregoing description of the embodiments of the present invention has been presented in conjunction with the drawings, it should be understood that it is not intended to limit the scope of the invention, but rather, it should be understood that various changes and modifications could be made by one skilled in the art without the need for inventive faculty, which would fall within the scope of the invention.
Claims (4)
1. The wearable dynamic electrocardiosignal classification method is characterized by comprising the following steps of:
acquiring an original electrocardiosignal to be classified;
Extracting wavelet scattering coefficients of electrocardiosignals to be classified by utilizing a pre-constructed wavelet scattering network, constructing a scattering feature matrix by utilizing the wavelet scattering coefficients, and converting the scattering feature matrix into a scattering feature map; three layers of wavelet scattering networks are constructed based on the scale function and the wavelet function, and 0-order scattering coefficients, 1-order scattering coefficients and 2-order scattering coefficients are generated;
processing the scattering feature map by using a pre-trained classification model to obtain a classification result of the electrocardiosignal, kicking out signals with serious noise pollution based on the classification result, and dividing the reserved signals into mild pollution signals only used for heart rate extraction and clean signals used for disease classification; the training set comprises a plurality of electrocardiosignals and labels for labeling the classes of the electrocardiosignals, and the training process of the model comprises the following steps:
Constructing a training set, wherein the training set is from a quality evaluation database, and the quality evaluation database comprises a plurality of electrocardiosignals with finished manual labeling, wherein the categories of the electrocardiosignals comprise clean signals for disease detection and diagnosis, mild pollution signals for heart rate extraction and noise pollution serious signals to be removed; all data in the quality assessment database are first preprocessed, including: data standardization processing, lead falling judgment and pure noise judgment;
Specifically, a long-time dynamic wearable dynamic electrocardiosignal under a set sampling frequency is obtained, the electrocardiosignal is segmented by utilizing a sliding window, the quality label of each limit number is determined, the problem of data imbalance is solved by adding noise, and a signal quality evaluation database is formed; conducting lead shedding judgment on the electrocardiosignals, regarding a group of ECG signals, if the number of samples with equal amplitude of two adjacent samples is larger than a set value of the total number of samples, considering the signals as lead shedding signals, and eliminating the signals; carrying out pure Gaussian noise judgment on the electrocardiosignal, and based on a set frequency spectrum range of the electrocardiosignal, if the ratio of the power spectrum energy of the signal in the set frequency spectrum range to the total energy is smaller than a preset value, considering the signal as a noise signal, and directly discarding the signal;
Taking the preprocessed signals as input signals of a three-layer wavelet scattering network, and calculating 0-order, 1-order and 2-order scattering coefficients; determining a maximum scale factor of which the scale function keeps translational invariance based on the length of the signal and the sampling frequency Fs, wherein the maximum scale factor is the real-time support degree I, and the value of the I is 2-10 xFs; convolving the input electrocardiosignals to be classified with a scale function to obtain a 0-order wavelet dispersion coefficient; the electrocardiosignals to be classified are convolved with a first-order wavelet function, and a first-order scattering propagation operator is generated through nonlinear modulo operation; convolving the first-order scattering propagation operator with the scale function to obtain a first-order wavelet scattering coefficient; the first-order scattering propagation operator is convolved with a second-order wavelet function, and a second-order scattering propagation operator is generated through nonlinear modulo operation; convolving the second-order scattering propagation operator with the scale function to obtain a second-order wavelet scattering coefficient;
And inputting the scattering feature matrix extracted by the training data wavelet scattering network into a classification model based on the long-short-term memory network LSTM, and training to obtain a trained classification model.
2. A wearable dynamic electrocardiograph signal classification system, comprising:
the acquisition module is used for acquiring the original electrocardiosignals to be classified;
The extraction module is used for extracting wavelet scattering coefficients of the electrocardiosignals to be classified by utilizing a pre-constructed wavelet scattering network, calculating a scattering feature matrix by utilizing the wavelet scattering coefficients and converting the scattering feature matrix into a scattering feature map; three layers of wavelet scattering networks are constructed based on the scale function and the wavelet function, and 0-order scattering coefficients, 1-order scattering coefficients and 2-order scattering coefficients are generated;
The classification module is used for processing the scattering feature map by utilizing a pre-trained classification model to obtain a classification result of the electrocardiosignal, kicking off a signal with serious noise pollution based on the classification result, and dividing a reserved signal into a mild pollution signal only used for heart rate extraction and a clean signal used for disease classification; the training set comprises a plurality of electrocardiosignals and labels for labeling the classes of the electrocardiosignals, and the training process of the model comprises the following steps:
Constructing a training set, wherein the training set is from a quality evaluation database, and the quality evaluation database comprises a plurality of electrocardiosignals with finished manual labeling, wherein the categories of the electrocardiosignals comprise clean signals for disease detection and diagnosis, mild pollution signals for heart rate extraction and noise pollution serious signals to be removed; all data in the quality assessment database are first preprocessed, including: data standardization processing, lead falling judgment and pure noise judgment;
Specifically, a long-time dynamic wearable dynamic electrocardiosignal under a set sampling frequency is obtained, the electrocardiosignal is segmented by utilizing a sliding window, the quality label of each limit number is determined, the problem of data imbalance is solved by adding noise, and a signal quality evaluation database is formed; conducting lead shedding judgment on the electrocardiosignals, regarding a group of ECG signals, if the number of samples with equal amplitude of two adjacent samples is larger than a set value of the total number of samples, considering the signals as lead shedding signals, and eliminating the signals; carrying out pure Gaussian noise judgment on the electrocardiosignal, and based on a set frequency spectrum range of the electrocardiosignal, if the ratio of the power spectrum energy of the signal in the set frequency spectrum range to the total energy is smaller than a preset value, considering the signal as a noise signal, and directly discarding the signal;
Taking the preprocessed signals as input signals of a three-layer wavelet scattering network, and calculating 0-order, 1-order and 2-order scattering coefficients; determining a maximum scale factor of which the scale function keeps translational invariance based on the length of the signal and the sampling frequency Fs, wherein the maximum scale factor is the real-time support degree I, and the value of the I is 2-10 xFs; convolving the input electrocardiosignals to be classified with a scale function to obtain a 0-order wavelet dispersion coefficient; the electrocardiosignals to be classified are convolved with a first-order wavelet function, and a first-order scattering propagation operator is generated through nonlinear modulo operation; convolving the first-order scattering propagation operator with the scale function to obtain a first-order wavelet scattering coefficient; the first-order scattering propagation operator is convolved with a second-order wavelet function, and a second-order scattering propagation operator is generated through nonlinear modulo operation; convolving the second-order scattering propagation operator with the scale function to obtain a second-order wavelet scattering coefficient;
And inputting the scattering feature matrix extracted by the training data wavelet scattering network into a classification model based on the long-short-term memory network LSTM, and training to obtain a trained classification model.
3. A non-transitory computer readable storage medium storing computer instructions that, when executed by a processor, implement the wearable dynamic electrocardiograph signal classification method of claim 1.
4. An electronic device, comprising: a processor, a memory, and a computer program; wherein the processor is connected to the memory, and the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory, so that the electronic device executes the instructions for implementing the wearable dynamic electrocardiograph signal classification method according to claim 1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111247597.6A CN113796873B (en) | 2021-10-26 | 2021-10-26 | Wearable dynamic electrocardiosignal classification method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111247597.6A CN113796873B (en) | 2021-10-26 | 2021-10-26 | Wearable dynamic electrocardiosignal classification method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113796873A CN113796873A (en) | 2021-12-17 |
| CN113796873B true CN113796873B (en) | 2024-05-03 |
Family
ID=78898296
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111247597.6A Active CN113796873B (en) | 2021-10-26 | 2021-10-26 | Wearable dynamic electrocardiosignal classification method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113796873B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115517685B (en) * | 2022-09-15 | 2024-11-05 | 山东大学 | Electrocardiosignal detection method and system based on dynamic learning and multi-scale analysis |
| CN116584902B (en) * | 2023-07-18 | 2024-05-07 | 山东大学 | Heart sound classification device based on feature optimization and visualization |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1427333A2 (en) * | 2001-09-21 | 2004-06-16 | Cardiomag Imaging | Nonlinear noise reduction for magnetocardiograms using wavelet transforms |
| WO2013036718A1 (en) * | 2011-09-08 | 2013-03-14 | Isis Innovation Ltd. | Determining acceptability of physiological signals |
| CN109117769A (en) * | 2018-07-31 | 2019-01-01 | 东南大学 | A kind of real-time quality assessment feedback method for wearing type electrocardiogram acquisition |
| WO2019075035A1 (en) * | 2017-10-12 | 2019-04-18 | Zoll Medical Corporation | System and method for distinguishing a cardiac event from noise in an electrocardiogram (ecg) signal |
| CN109745033A (en) * | 2018-12-25 | 2019-05-14 | 东南大学 | Dynamic electrocardiogram method for evaluating quality based on time-frequency two-dimensional image and machine learning |
| WO2019100565A1 (en) * | 2017-11-27 | 2019-05-31 | 乐普（北京）医疗器械股份有限公司 | Method and device for self-learning dynamic electrocardiography analysis employing artificial intelligence |
| CN111160139A (en) * | 2019-12-13 | 2020-05-15 | 中国科学院深圳先进技术研究院 | Electrocardiosignal processing method and device and terminal equipment |
| CN111956221A (en) * | 2020-09-07 | 2020-11-20 | 南京医科大学 | Temporal lobe epilepsy classification method based on wavelet scattering factor and LSTM neural network model |
| CN113343805A (en) * | 2021-05-26 | 2021-09-03 | 南京医科大学 | Atrial fibrillation rhythm identification method based on RR interphase electrocardiogram data and ensemble learning |
| CN113361385A (en) * | 2021-06-03 | 2021-09-07 | 山东大学 | Heart sound classification method and system, readable storage medium and electronic device |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060079773A1 (en) * | 2000-11-28 | 2006-04-13 | Allez Physionix Limited | Systems and methods for making non-invasive physiological assessments by detecting induced acoustic emissions |
| US7022077B2 (en) * | 2000-11-28 | 2006-04-04 | Allez Physionix Ltd. | Systems and methods for making noninvasive assessments of cardiac tissue and parameters |
| WO2020136570A1 (en) * | 2018-12-26 | 2020-07-02 | Analytics For Life Inc. | Method and system for automated quantification of signal quality |
-
2021
- 2021-10-26 CN CN202111247597.6A patent/CN113796873B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1427333A2 (en) * | 2001-09-21 | 2004-06-16 | Cardiomag Imaging | Nonlinear noise reduction for magnetocardiograms using wavelet transforms |
| WO2013036718A1 (en) * | 2011-09-08 | 2013-03-14 | Isis Innovation Ltd. | Determining acceptability of physiological signals |
| WO2019075035A1 (en) * | 2017-10-12 | 2019-04-18 | Zoll Medical Corporation | System and method for distinguishing a cardiac event from noise in an electrocardiogram (ecg) signal |
| WO2019100565A1 (en) * | 2017-11-27 | 2019-05-31 | 乐普（北京）医疗器械股份有限公司 | Method and device for self-learning dynamic electrocardiography analysis employing artificial intelligence |
| CN109117769A (en) * | 2018-07-31 | 2019-01-01 | 东南大学 | A kind of real-time quality assessment feedback method for wearing type electrocardiogram acquisition |
| CN109745033A (en) * | 2018-12-25 | 2019-05-14 | 东南大学 | Dynamic electrocardiogram method for evaluating quality based on time-frequency two-dimensional image and machine learning |
| CN111160139A (en) * | 2019-12-13 | 2020-05-15 | 中国科学院深圳先进技术研究院 | Electrocardiosignal processing method and device and terminal equipment |
| CN111956221A (en) * | 2020-09-07 | 2020-11-20 | 南京医科大学 | Temporal lobe epilepsy classification method based on wavelet scattering factor and LSTM neural network model |
| CN113343805A (en) * | 2021-05-26 | 2021-09-03 | 南京医科大学 | Atrial fibrillation rhythm identification method based on RR interphase electrocardiogram data and ensemble learning |
| CN113361385A (en) * | 2021-06-03 | 2021-09-07 | 山东大学 | Heart sound classification method and system, readable storage medium and electronic device |
Non-Patent Citations (3)
| Title |
|---|
| 《Wavelet Scattering Transform for ECG Beat Classification》;Liu Z,Yao G,Zhang J,Zeng X;《Comput Math Methods Med》;第7卷;正文1-10页 * |
| Liu Z,Yao G,Zhang J,Zeng X.《Wavelet Scattering Transform for ECG Beat Classification》.《Comput Math Methods Med》.2020,第7卷正文1-10页. * |
| 基于脑电（EEG）的帕金森疾病早期诊断的研究;骆春桃;中国优秀硕士学位论文全文数据库;20200615(第06期);E070-30 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113796873A (en) | 2021-12-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Übeyli | Combined neural network model employing wavelet coefficients for EEG signals classification | |
| Sridhar et al. | Brain tumor classification using discrete cosine transform and probabilistic neural network | |
| CN105956623A (en) | Epilepsy electroencephalogram signal classification method based on fuzzy entropy | |
| CN113796873B (en) | Wearable dynamic electrocardiosignal classification method and system | |
| Ravi et al. | Epileptic seizure detection using convolutional neural networks and recurrence plots of EEG signals | |
| Mouleeshuwarapprabu et al. | Nonlinear vector decomposed neural network based EEG signal feature extraction and detection of seizure | |
| Qin et al. | Deep multi-scale feature fusion convolutional neural network for automatic epilepsy detection using EEG signals | |
| CN113076878A (en) | Physique identification method based on attention mechanism convolution network structure | |
| CN114464319B (en) | AMS susceptibility assessment system based on slow feature analysis and deep neural network | |
| Sadiq et al. | Auto-correlation based feature extraction approach for EEG alcoholism identification | |
| Singh et al. | An intelligent method for epilepsy seizure detection based on hybrid nonlinear EEG data features using adaptive signal decomposition methods | |
| AlSharabi et al. | EEG-based clinical decision support system for Alzheimer's disorders diagnosis using EMD and deep learning techniques | |
| Hasanah et al. | Classification of brain tumor on magnetic resonance imaging using support vector machine | |
| CN112603337A (en) | Electroencephalogram signal identification method | |
| CN116898454B (en) | Epileptic classification method and system based on electroencephalogram feature fusion deep learning model | |
| Maleki et al. | Identification of cardiovascular diseases through ECG classification using wavelet transformation | |
| CN116491956A (en) | Abnormal electroencephalogram automatic detection method based on multi-domain feature fusion | |
| Islam et al. | Genetic algorithm based optimal feature selection extracted by time-frequency analysis for enhanced sleep disorder diagnosis using eeg signal | |
| Wen et al. | Intelligent personalized diagnosis modeling in advanced medical system for Parkinson’s disease using voice signals | |
| Satapathy et al. | A deep neural model CNN-LSTM network for automated sleep staging based on a single-channel EEG signal | |
| CN113558643A (en) | Multi-feature epilepsy signal classification method based on VMD and NLTWSVM | |
| Sun et al. | Multi-Granularity Graph Convolution Network for Major Depressive Disorder Recognition | |
| Sasikala et al. | Prediction of heart stroke diseases using machine learning technique based electromyographic data | |
| Al-Qazzaz et al. | Automatic Diagnosis of Epileptic Seizures using Entropy-based Features and Multimodel Deep Learning Approaches | |
| CN118418738B (en) | Method and device for evaluating health state of battery of electric automobile |
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 |