CN112446307A - Local constraint-based non-negative matrix factorization electrocardiogram identity recognition method and system - Google Patents
Local constraint-based non-negative matrix factorization electrocardiogram identity recognition method and system Download PDFInfo
- Publication number
- CN112446307A CN112446307A CN202011278357.8A CN202011278357A CN112446307A CN 112446307 A CN112446307 A CN 112446307A CN 202011278357 A CN202011278357 A CN 202011278357A CN 112446307 A CN112446307 A CN 112446307A
- Authority
- CN
- China
- Prior art keywords
- matrix
- electrocardiosignal
- electrocardiosignals
- monocycle
- identified
- 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.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/02—Preprocessing
- G06F2218/04—Denoising
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Signal Processing (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention discloses a local constraint based non-negative matrix factorization electrocardio identity recognition method and a system, comprising the following steps: acquiring an electrocardiosignal to be identified; performing heart beat division on the electrocardiosignals to be identified to obtain a plurality of monocycle electrocardiosignals; for each monocycle electrocardiosignal, projecting by using a pre-obtained pseudo-inverse matrix of the basis matrix, and projecting the electrocardiosignal from a high-dimensional space to a low-dimensional space to obtain a characteristic representation after dimension reduction; the method comprises the steps that a pseudo-inverse matrix of a base matrix which is obtained in advance is obtained by processing a electrocardiosignal training data set in a non-negative matrix decomposition mode based on local constraint; and inputting the characteristic representation after dimension reduction into the trained classification model to obtain an identity recognition result corresponding to the current electrocardiosignal to be recognized.
Description
Technical Field
The application relates to the technical field of identity recognition, in particular to an electrocardio identity recognition method and system based on local constraint and nonnegative matrix factorization.
Background
The statements in this section merely provide background information related to the present disclosure and may not constitute prior art.
In recent years, the universality and the distinguishability of electrocardiosignals are widely concerned and applied to identity recognition. The electrocardiosignals are important physiological signals on human bodies, and have difference among different individuals, the electrocardiosignals cannot change greatly within a period of time, and the collection of the electrocardiosignals is more convenient along with the development of the micro sensor technology. Therefore, the electrocardiosignal has the characteristics of universality, identifiability, stability, easy acquisition and the like, meets the premise of biological characteristic identification, is a safer and more reliable identity identification technology, and has good application prospect. However, although the identification technology based on the electrocardiosignal has some advantages, the existing identification technology still has a plurality of problems to be overcome. How to better remove interference information and extract an electrocardio identity recognition characteristic with high discriminability is one of the problems to be solved urgently at present.
Disclosure of Invention
In order to overcome the defects of the prior art, the application provides a local constraint-based non-negative matrix factorization electrocardiogram identity recognition method and system; according to the method, according to an electrocardiosignal structure, filtering, heart beat division and the like are carried out on electrocardiosignal waveforms, a local non-negative matrix decomposition (LNMF) method is adopted to obtain a base matrix W and a coefficient matrix H, and any electrocardiosignal sample can be projected from a high-dimensional space to a low-dimensional space through a pseudo-inverse matrix of the base matrix to obtain an extracted feature matrix. And then classifying the electrocardiosignals after the characteristics are extracted by using a Relevance Vector Machine (RVM) so as to improve the performance and the robustness of the electrocardiosignal identity identification method.
In a first aspect, the application provides a local constraint-based non-negative matrix factorization electrocardiogram identity recognition method;
the electrocardio identity recognition method based on the local constraint non-negative matrix factorization comprises the following steps:
acquiring an electrocardiosignal to be identified;
performing heart beat division on the electrocardiosignals to be identified to obtain a plurality of monocycle electrocardiosignals;
for each monocycle electrocardiosignal, projecting by using a pre-obtained pseudo-inverse matrix of the basis matrix, and projecting the electrocardiosignal from a high-dimensional space to a low-dimensional space to obtain a characteristic representation after dimension reduction;
the method comprises the steps that a pseudo-inverse matrix of a base matrix which is obtained in advance is obtained by processing a electrocardiosignal training data set in a non-negative matrix decomposition mode based on local constraint;
and inputting the characteristic representation after dimension reduction into the trained classification model to obtain an identity recognition result corresponding to the current electrocardiosignal to be recognized.
In a second aspect, the application provides a local constraint-based non-negative matrix factorization electrocardiogram identity recognition system;
the electrocardio identity recognition system based on local constraint non-negative matrix factorization comprises:
an acquisition module configured to: acquiring an electrocardiosignal to be identified;
a partitioning module configured to: performing heart beat division on the electrocardiosignals to be identified to obtain a plurality of monocycle electrocardiosignals;
a projection module configured to: for each monocycle electrocardiosignal, projecting by using a pre-obtained pseudo-inverse matrix of the basis matrix, and projecting the electrocardiosignal from a high-dimensional space to a low-dimensional space to obtain a characteristic representation after dimension reduction;
the method comprises the steps that a pseudo-inverse matrix of a base matrix which is obtained in advance is obtained by processing a electrocardiosignal training data set in a non-negative matrix decomposition mode based on local constraint;
an identification module configured to: and inputting the characteristic representation after dimension reduction into the trained classification model to obtain an identity recognition result corresponding to the current electrocardiosignal to be recognized.
In a third aspect, the present application further provides an electronic device, including: one or more processors, one or more memories, and one or more computer programs; wherein a processor is connected to the memory, the one or more computer programs are stored in the memory, and when the electronic device is running, the processor executes the one or more computer programs stored in the memory, so as to make the electronic device execute the method according to the first aspect.
In a fourth aspect, the present application also provides a computer-readable storage medium for storing computer instructions which, when executed by a processor, perform the method of the first aspect.
In a fifth aspect, the present application also provides a computer program (product) comprising a computer program for implementing the method of any of the preceding first aspects when run on one or more processors.
Compared with the prior art, the beneficial effects of this application are:
the method has the advantages that through the preprocessing processes of denoising, filtering and the like on the electrocardiosignals, the noise which has large influence on the electrocardiosignals is removed, and the purer electrocardiosignals are obtained; and decomposing the divided heart beats through a local non-negative matrix to obtain a base matrix W and a coefficient matrix H, calculating to obtain a pseudo-inverse matrix M of the base matrix, and obtaining the mapped electrocardiosignals by utilizing the M. LNMF is a significant improvement over traditional NMF algorithms, emphasizes feature locality in the base component, and defines an objective function to impose local constraints as well as non-negative constraints in order to render the form suitable for the task of feature localization. Therefore, the capability of capturing local features is higher than that of the traditional NMF, and the algorithm is more robust as a whole. The low-dimensional signals obtained by the LNMF algorithm have more detailed information and can better reflect the essential characteristics of the heartbeat. And finally, sending the mapped low-dimensional signals into a related vector machine with excellent performance on multi-classification problems for classification, thereby improving the identification effect of the electrocardiosignals.
Advantages of additional aspects 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
The accompanying drawings, which are incorporated in and constitute a part of this application, illustrate embodiments of the application and, together with the description, serve to explain the application and are not intended to limit the application.
FIG. 1 is a flow chart of the method of the first embodiment.
Detailed Description
It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the disclosure. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of example embodiments according to the present application. As used herein, the singular forms "a", "an", and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise, and it should be understood that the terms "comprises" and "comprising", and any variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
The embodiments and features of the embodiments of the present invention may be combined with each other without conflict.
Example one
The embodiment provides an electrocardio identity recognition method based on local constraint non-negative matrix factorization;
as shown in fig. 1, the local constraint-based electrocardio identity recognition method based on non-negative matrix factorization includes:
s101: acquiring an electrocardiosignal to be identified;
s102: performing heart beat division on the electrocardiosignals to be identified to obtain a plurality of monocycle electrocardiosignals;
s103: for each monocycle electrocardiosignal, projecting by using a pre-obtained pseudo-inverse matrix of the basis matrix, and projecting the electrocardiosignal from a high-dimensional space to a low-dimensional space to obtain a characteristic representation after dimension reduction;
the method comprises the steps that a pseudo-inverse matrix of a base matrix which is obtained in advance is obtained by processing a electrocardiosignal training data set in a non-negative matrix decomposition mode based on local constraint;
s104: and inputting the characteristic representation after dimension reduction into the trained classification model to obtain an identity recognition result corresponding to the current electrocardiosignal to be recognized.
As one or more embodiments, after the step of S101 acquiring an electrocardiographic signal to be recognized, before the step of S102 performing cardiac beat division on the electrocardiographic signal to be recognized to obtain a plurality of single-cycle electrocardiographic signals, the method further includes:
preprocessing the electrocardiosignals to be identified.
Further, the preprocessing the electrocardiographic signal to be recognized specifically includes:
denoising and filtering the electrocardiosignals to be identified.
Illustratively, some noise interference exists in the original electrocardiosignal, such as myoelectric interference, power frequency interference, baseline drift and the like. A normal electrocardiosignal frequency range is 0.5-100Hz, wherein 1-40Hz contains most energy of the electrocardiosignal, the amplitude of the electrocardiosignal is weak, and excessive noise can seriously affect the waveform and the amplitude of the electrocardiosignal. According to the scheme, a 4-order Butterworth band-pass filter is adopted mainly according to the frequency range of electrocardiosignals, the low-frequency cut-off frequency of the filter is set to be 1Hz, and the high-frequency cut-off frequency of the filter is set to be 40 Hz.
As one or more embodiments, in S102, cardiac beat division is performed on the electrocardiographic signals to be identified to obtain a plurality of monocycle electrocardiographic signals; the method comprises the following specific steps:
positioning an R peak through a pan _ tompkin algorithm, and taking a set number of points forwards and backwards according to the sampling frequency of the electrocardiosignals by taking the R peak as a reference to form a heart beat, wherein the heart beat is the monocycle electrocardiosignals;
for each electrocardiosignal to be identified, dividing a plurality of monocycle electrocardiosignals by using a pan _ tempkin algorithm, and calculating average heart beat;
calculating the difference between each monocycle electrocardiosignal and the average heart beat;
and comparing the difference with a set threshold, eliminating the monocycle electrocardiosignals of which the difference is greater than the set threshold, and reserving the monocycle electrocardiosignals of which the difference is less than the set threshold.
Illustratively, the electrocardiosignal is a non-stationary periodic-like signal and consists of a plurality of heart beats, and a complete heart beat mainly comprises a P wave band, a QRS wave band and a T wave band. In the mapping of the cardiac signal waveform, the R peak is the most prominent peak and can be used as a marker for a given heartbeat waveform. And positioning the R peak through a pan _ tompkin algorithm, and taking a certain number of points forwards and backwards by taking the R peak as a reference according to the sampling frequency of the electrocardiosignal to form a heart beat. And segmenting each electrocardiogram record into a plurality of heart beats by using a pan _ tompkin algorithm, calculating an average heart beat, and screening out the heart beats which are too far away from the average heart beat by setting a threshold value.
As will be appreciated, the difference between each monocycle cardiac signal and the average beat is calculated; comparing the difference with a set threshold, eliminating the monocycle electrocardiosignals of which the difference is greater than the set threshold, and reserving the monocycle electrocardiosignals of which the difference is less than the set threshold; the advantage of processing in this way is that the interference of irregular monocycle electrocardiosignals to the final identity recognition result can be avoided, and the preservation of regular monocycle electrocardiosignals has great significance for improving the accuracy of the identity recognition result.
As one or more embodiments, in S103, the pre-obtained pseudo-inverse matrix of the basis matrix is obtained by processing the electrocardiosignal training data set in a non-negative matrix decomposition manner based on local constraint; the method comprises the following specific steps:
s1031: carrying out heart beat division processing on the electrocardiosignals in the training set to obtain a plurality of monocycle electrocardiosignals;
s1032: each monocycle electrocardiosignal is translated upwards by p units to obtain a non-negative monocycle electrocardiosignal; p is a positive integer;
s1033: taking the amplitude of each moment of each non-negative monocycle electrocardiosignal as one column of a matrix, and forming an m x n matrix V by n m-dimensional non-negative monocycle electrocardiosignals; n and m are positive integers;
s1033: executing a non-negative matrix decomposition algorithm based on local constraint on the matrix V to obtain a base matrix W and a coefficient matrix H; and obtaining a pseudo-inverse matrix M of the base matrix W based on the base matrix W.
Illustratively, the heartbeat signal has positive and negative values, and in order to meet the requirement of the algorithm on the premise of ensuring that the geometric structure is not changed, the heartbeat signal is firstly translated upwards by p (n >0) units, so that the non-negative heartbeat signal is obtained. Taking each non-negative heart beat signal as a column of the matrix, and forming an m x n matrix V by n m-dimensional non-negative heart beat signals; after a local non-negative matrix decomposition algorithm is executed on the matrix V, a base matrix W and a coefficient matrix H can be obtained, and a pseudo-inverse matrix M of the base matrix W is obtained through calculation; and mapping the non-negative heart beat by using the learned M matrix to obtain the low-dimensional characteristic representation of the signal.
Let A be [ a ═ aij]＝WTW,B＝[bij]＝HHTThe purpose of the LNMF is to learn the features, Σ, by imposing three additional constraints on the NMF basisiaii＝min,∑i≠jaij＝min,∑ibii＝maxThe objective function is as follows:
wherein, alpha, beta is more than 0, and has regulating effect.
The solving process by the objective function local non-negative matrix decomposition is as follows:
inputting: initial non-negative matrix V, reduced dimension data dimension k
And (3) outputting: base matrix W and coefficient matrix H
step1 random initialization of positive value matrices W and H, setting the maximum number of iterations nmax
Step2:for count＝1：nmax do
Step3, sequentially iterating all the elements in W and H according to the following rules:
Step4:end for
and (3) calculating a pseudo-inverse matrix M of the W according to a formula (2) from the solved W:
M＝(WTW)-1WT (2)
the original non-negative heart beat signal can be mapped by M, and the electrocardiosignal x of M dimensioniNamely, k-dimensional signal y can be obtained by mapping formula (3)i：
yi＝Mxi (3)
As one or more embodiments, in S104, the feature representation after the dimension reduction is input into the trained classification model to obtain an identity recognition result corresponding to the current electrocardiographic signal to be recognized; the training step of the specifically trained classification model comprises the following steps:
constructing a classification model;
constructing a training set; the training set is electrocardio characteristic representation of a known identity recognition result;
and inputting the training set into the classification model, and stopping training when the loss function of the classification model reaches the minimum value to obtain the trained classification model.
Further, the classification model selects an RVM classifier.
Further, the step of obtaining the training set comprises:
acquiring electrocardiosignals of known identity types;
preprocessing the electrocardiosignal;
performing heart beat division on the preprocessed electrocardiosignals to obtain a plurality of monocycle electrocardiosignals;
for each single-cycle electrocardiosignal, obtaining a basis matrix based on a non-negative matrix decomposition algorithm of local constraint;
obtaining a pseudo-inverse matrix of the basis matrix based on the basis of the basis matrix;
and mapping the non-negative monocycle electrocardiosignals based on the pseudo-inverse matrix of the basis matrix to obtain the electrocardio characteristic representation.
For a new input heartbeat x to be recognizediMapping by using the pseudo-inverse matrix M of the basis matrix W learned by the training set V and the formula (3) to obtain the low-dimensional expression yiAnd the dimensionality reduction process is also completed for the heart beat signal of the training set V to obtain a low-dimensional representation.
In the identification and classification process, the RVM model is trained by using the low-dimensional representation of the training set V, and the appropriate parameters are adjusted and selected. Then the low-dimensional representation of the heart beat y to be recognized is takeniAnd sending the model into a trained RVM model for classification to obtain a final recognition result.
The RVM is proposed on the basis of a Bayes framework, a low-dimensional space nonlinear problem can be converted into a high-dimensional space linear problem on the basis of kernel function mapping, the multi-classification problem processing base is firmer, and a probability form prediction can be given to a new data point.
Example two
The embodiment provides an electrocardio identity recognition system based on local constraint and nonnegative matrix factorization;
the electrocardio identity recognition system based on local constraint non-negative matrix factorization comprises:
an acquisition module configured to: acquiring an electrocardiosignal to be identified;
a partitioning module configured to: performing heart beat division on the electrocardiosignals to be identified to obtain a plurality of monocycle electrocardiosignals;
a projection module configured to: for each monocycle electrocardiosignal, projecting by using a pre-obtained pseudo-inverse matrix of the basis matrix, and projecting the electrocardiosignal from a high-dimensional space to a low-dimensional space to obtain a characteristic representation after dimension reduction;
the method comprises the steps that a pseudo-inverse matrix of a base matrix which is obtained in advance is obtained by processing a electrocardiosignal training data set in a non-negative matrix decomposition mode based on local constraint;
an identification module configured to: and inputting the characteristic representation after dimension reduction into the trained classification model to obtain an identity recognition result corresponding to the current electrocardiosignal to be recognized.
It should be noted here that the acquiring module, the dividing module, the projecting module and the identifying module correspond to steps S101 to S104 in the first embodiment, and the modules are the same as the examples and application scenarios realized by the corresponding steps, but are not limited to the disclosure of the first embodiment. It should be noted that the modules described above as part of a system may be implemented in a computer system such as a set of computer-executable instructions.
In the foregoing embodiments, the descriptions of the embodiments have different emphasis, and for parts that are not described in detail in a certain embodiment, reference may be made to related descriptions of other embodiments.
The proposed system can be implemented in other ways. For example, the above-described system embodiments are merely illustrative, and for example, the division of the above-described modules is merely a logical division, and in actual implementation, there may be other divisions, for example, multiple modules may be combined or integrated into another system, or some features may be omitted, or not executed.
EXAMPLE III
The present embodiment also provides an electronic device, including: one or more processors, one or more memories, and one or more computer programs; wherein, a processor is connected with the memory, the one or more computer programs are stored in the memory, and when the electronic device runs, the processor executes the one or more computer programs stored in the memory, so as to make the electronic device execute the method according to the first embodiment.
It should be understood that in this embodiment, the processor may be a central processing unit CPU, and the processor may also be other general purpose processors, digital signal processors DSP, application specific integrated circuits ASIC, off-the-shelf programmable gate arrays FPGA or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, and so on. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
The memory may include both read-only memory and random access memory, and may provide instructions and data to the processor, and a portion of the memory may also include non-volatile random access memory. For example, the memory may also store device type information.
In implementation, the steps of the above method may be performed by integrated logic circuits of hardware in a processor or instructions in the form of software.
The method in the first embodiment may be directly implemented by a hardware processor, or may be implemented by a combination of hardware and software modules in the processor. The software modules may be located in ram, flash, rom, prom, or eprom, registers, among other storage media as is well known in the art. The storage medium is located in a memory, and a processor reads information in the memory and completes the steps of the method in combination with hardware of the processor. To avoid repetition, it is not described in detail here.
Those of ordinary skill in the art will appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.
Example four
The present embodiments also provide a computer-readable storage medium for storing computer instructions, which when executed by a processor, perform the method of the first embodiment.
The above description is only a preferred embodiment of the present application and is not intended to limit the present application, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, improvement and the like made within the spirit and principle of the present application shall be included in the protection scope of the present application.
Claims (10)
1. The electrocardio identity recognition method based on local constraint non-negative matrix decomposition is characterized by comprising the following steps:
acquiring an electrocardiosignal to be identified;
performing heart beat division on the electrocardiosignals to be identified to obtain a plurality of monocycle electrocardiosignals;
for each monocycle electrocardiosignal, projecting by using a pre-obtained pseudo-inverse matrix of the basis matrix, and projecting the electrocardiosignal from a high-dimensional space to a low-dimensional space to obtain a characteristic representation after dimension reduction;
the method comprises the steps that a pseudo-inverse matrix of a base matrix which is obtained in advance is obtained by processing a electrocardiosignal training data set in a non-negative matrix decomposition mode based on local constraint;
and inputting the characteristic representation after dimension reduction into the trained classification model to obtain an identity recognition result corresponding to the current electrocardiosignal to be recognized.
2. The method as claimed in claim 1, wherein after the step of obtaining the cardiac signal to be identified, the step of dividing the cardiac signal to be identified into a plurality of monocycle cardiac signals further comprises, before the step of obtaining the plurality of monocycle cardiac signals:
preprocessing the electrocardiosignals to be identified.
3. The method according to claim 2, wherein the preprocessing of the cardiac signal to be identified comprises:
denoising and filtering the electrocardiosignals to be identified.
4. The method according to claim 1, characterized in that the cardiac electrical signal to be identified is subjected to a cardiac beat division, obtaining a plurality of monocycle cardiac electrical signals; the method comprises the following specific steps:
positioning an R peak through a pan _ tompkin algorithm, and taking a set number of points forwards and backwards according to the sampling frequency of the electrocardiosignals by taking the R peak as a reference to form a heart beat, wherein the heart beat is the monocycle electrocardiosignals;
for each electrocardiosignal to be identified, dividing a plurality of monocycle electrocardiosignals by using a pan _ tempkin algorithm, and calculating average heart beat;
calculating the difference between each monocycle electrocardiosignal and the average heart beat;
and comparing the difference with a set threshold, eliminating the monocycle electrocardiosignals of which the difference is greater than the set threshold, and reserving the monocycle electrocardiosignals of which the difference is less than the set threshold.
5. The method of claim 1, wherein the pre-obtained pseudo-inverse of the basis matrix is obtained by processing the training data set of the electrocardiosignal in a non-negative matrix decomposition based on local constraints; the method comprises the following specific steps:
carrying out heart beat division processing on the electrocardiosignals in the training set to obtain a plurality of monocycle electrocardiosignals;
each monocycle electrocardiosignal is translated upwards by p units to obtain a non-negative monocycle electrocardiosignal; p is a positive integer;
taking the amplitude of each moment of each non-negative monocycle electrocardiosignal as one column of a matrix, and forming an m x n matrix V by n m-dimensional non-negative monocycle electrocardiosignals; n and m are positive integers;
executing a non-negative matrix decomposition algorithm based on local constraint on the matrix V to obtain a base matrix W and a coefficient matrix H; and obtaining a pseudo-inverse matrix M of the base matrix W based on the base matrix W.
6. The method as claimed in claim 1, wherein the feature representation after dimension reduction is input into the trained classification model to obtain the identity recognition result corresponding to the current electrocardiosignal to be recognized; the training step of the specifically trained classification model comprises the following steps:
constructing a classification model;
constructing a training set; the training set is electrocardio characteristic representation of a known identity recognition result;
and inputting the training set into the classification model, and stopping training when the loss function of the classification model reaches the minimum value to obtain the trained classification model.
7. The method of claim 6, wherein the training set obtaining step comprises:
acquiring electrocardiosignals of known identity types;
preprocessing the electrocardiosignal;
performing heart beat division on the preprocessed electrocardiosignals to obtain a plurality of monocycle electrocardiosignals;
for each single-cycle electrocardiosignal, obtaining a basis matrix based on a non-negative matrix decomposition algorithm of local constraint;
obtaining a pseudo-inverse matrix of the basis matrix based on the basis of the basis matrix;
and mapping the non-negative monocycle electrocardiosignals based on the pseudo-inverse matrix of the basis matrix to obtain the electrocardio characteristic representation.
8. The electrocardio identity recognition system based on local constraint and nonnegative matrix decomposition is characterized by comprising the following components:
an acquisition module configured to: acquiring an electrocardiosignal to be identified;
a partitioning module configured to: performing heart beat division on the electrocardiosignals to be identified to obtain a plurality of monocycle electrocardiosignals;
a projection module configured to: for each monocycle electrocardiosignal, projecting by using a pre-obtained pseudo-inverse matrix of the basis matrix, and projecting the electrocardiosignal from a high-dimensional space to a low-dimensional space to obtain a characteristic representation after dimension reduction;
the method comprises the steps that a pseudo-inverse matrix of a base matrix which is obtained in advance is obtained by processing a electrocardiosignal training data set in a non-negative matrix decomposition mode based on local constraint;
an identification module configured to: and inputting the characteristic representation after dimension reduction into the trained classification model to obtain an identity recognition result corresponding to the current electrocardiosignal to be recognized.
9. An electronic device, comprising: one or more processors, one or more memories, and one or more computer programs; wherein a processor is connected to the memory, the one or more computer programs being stored in the memory, the processor executing the one or more computer programs stored in the memory when the electronic device is running, to cause the electronic device to perform the method of any of the preceding claims 1-7.
10. A computer-readable storage medium storing computer instructions which, when executed by a processor, perform the method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011278357.8A CN112446307B (en) | 2020-11-16 | 2020-11-16 | Local constraint-based non-negative matrix factorization electrocardiogram identity recognition method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011278357.8A CN112446307B (en) | 2020-11-16 | 2020-11-16 | Local constraint-based non-negative matrix factorization electrocardiogram identity recognition method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112446307A true CN112446307A (en) | 2021-03-05 |
| CN112446307B CN112446307B (en) | 2022-11-04 |
Family
ID=74738464
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011278357.8A Active CN112446307B (en) | 2020-11-16 | 2020-11-16 | Local constraint-based non-negative matrix factorization electrocardiogram identity recognition method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112446307B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113221736A (en) * | 2021-05-11 | 2021-08-06 | 菏泽学院 | Electrocardio identity recognition method and system based on multi-feature collaborative non-negative matrix factorization |
| CN114325253A (en) * | 2021-11-22 | 2022-04-12 | 国网山东省电力公司电力科学研究院 | Partial discharge identification method and system for gas insulated metal enclosed switchgear |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107122643A (en) * | 2017-04-07 | 2017-09-01 | 西安电子科技大学 | Personal identification method based on PPG signals and breath signal Fusion Features |
| CN107239684A (en) * | 2017-05-22 | 2017-10-10 | 吉林大学 | A kind of feature learning method and system for ECG identifications |
| CN108537100A (en) * | 2017-11-17 | 2018-09-14 | 吉林大学 | A kind of electrocardiosignal personal identification method and system based on PCA and LDA analyses |
| CN109330585A (en) * | 2018-11-16 | 2019-02-15 | 山东大学 | A kind of electrocardiosignal identification, evaluation method and device based on NMF |
| CN110334761A (en) * | 2019-07-03 | 2019-10-15 | 北京林业大学 | There is supervision image-recognizing method based on Orthonormality constraints increment Non-negative Matrix Factorization |
| CN110786850A (en) * | 2019-11-11 | 2020-02-14 | 菏泽学院 | Electrocardiosignal identity recognition method and system based on multi-feature sparse representation |
| CN110786849A (en) * | 2019-11-11 | 2020-02-14 | 菏泽学院 | Electrocardiosignal identity recognition method and system based on multi-view discriminant analysis |
-
2020
- 2020-11-16 CN CN202011278357.8A patent/CN112446307B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107122643A (en) * | 2017-04-07 | 2017-09-01 | 西安电子科技大学 | Personal identification method based on PPG signals and breath signal Fusion Features |
| CN107239684A (en) * | 2017-05-22 | 2017-10-10 | 吉林大学 | A kind of feature learning method and system for ECG identifications |
| CN108537100A (en) * | 2017-11-17 | 2018-09-14 | 吉林大学 | A kind of electrocardiosignal personal identification method and system based on PCA and LDA analyses |
| CN109330585A (en) * | 2018-11-16 | 2019-02-15 | 山东大学 | A kind of electrocardiosignal identification, evaluation method and device based on NMF |
| CN110334761A (en) * | 2019-07-03 | 2019-10-15 | 北京林业大学 | There is supervision image-recognizing method based on Orthonormality constraints increment Non-negative Matrix Factorization |
| CN110786850A (en) * | 2019-11-11 | 2020-02-14 | 菏泽学院 | Electrocardiosignal identity recognition method and system based on multi-feature sparse representation |
| CN110786849A (en) * | 2019-11-11 | 2020-02-14 | 菏泽学院 | Electrocardiosignal identity recognition method and system based on multi-view discriminant analysis |
Non-Patent Citations (2)
| Title |
|---|
| MING TONG等: "NMF with local constraint and Deep NMF with temporal dependencies constraint for action recognition", 《NEURAL COMPUTING AND APPLICATIONS》 * |
| 王洋: "基于局部约束的非负矩阵分解方法研究", 《中国优秀博硕士学位论文全文数据库(硕士) 信息科技辑》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113221736A (en) * | 2021-05-11 | 2021-08-06 | 菏泽学院 | Electrocardio identity recognition method and system based on multi-feature collaborative non-negative matrix factorization |
| CN114325253A (en) * | 2021-11-22 | 2022-04-12 | 国网山东省电力公司电力科学研究院 | Partial discharge identification method and system for gas insulated metal enclosed switchgear |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112446307B (en) | 2022-11-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhang et al. | Automated detection of myocardial infarction using a gramian angular field and principal component analysis network | |
| Zhao et al. | ECG authentication system design incorporating a convolutional neural network and generalized S-Transformation | |
| Yildirim | A novel wavelet sequence based on deep bidirectional LSTM network model for ECG signal classification | |
| CN114052735B (en) | Deep field self-adaption-based electroencephalogram emotion recognition method and system | |
| Zhao et al. | ECG classification using deep CNN improved by wavelet transform | |
| Orhan | Real-time CHF detection from ECG signals using a novel discretization method | |
| CN111797674B (en) | MI electroencephalogram signal identification method based on feature fusion and particle swarm optimization algorithm | |
| CN107239684A (en) | A kind of feature learning method and system for ECG identifications | |
| CN108742697B (en) | Heart sound signal classification method and terminal equipment | |
| Zhang et al. | ECG-based personal recognition using a convolutional neural network | |
| CN111202512A (en) | Electrocardiogram classification method and device based on wavelet transformation and DCNN | |
| CN112426160A (en) | Electrocardiosignal type identification method and device | |
| CN112603330A (en) | Electrocardiosignal identification and classification method | |
| CN112446307B (en) | Local constraint-based non-negative matrix factorization electrocardiogram identity recognition method and system | |
| CN110786849B (en) | Electrocardiosignal identity recognition method and system based on multi-view discriminant analysis | |
| Zhao et al. | ECG identification based on matching pursuit | |
| Huang et al. | Robust multi-feature collective non-negative matrix factorization for ECG biometrics | |
| CN114052744A (en) | Electrocardiosignal classification method based on pulse neural network | |
| WO2018120088A1 (en) | Method and apparatus for generating emotional recognition model | |
| CN113569997A (en) | Emotion classification method and system based on graph convolution neural network | |
| CN107045624B (en) | Electroencephalogram signal preprocessing and classifying method based on maximum weighted cluster | |
| CN115281676A (en) | Fatigue detection method based on GRU neural network and ECG signal | |
| CN114947850A (en) | Mental load grade objective detection method based on pulse Bouss model characteristics | |
| CN111449646B (en) | Electrocardiogram identification method and device, storage medium and electronic equipment | |
| CN114692682A (en) | Method and system for classifying motor imagery based on graph embedding representation |
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 |