ECG signal classification system, method, equipment, medium and product

Abstract

The invention discloses an ECG signal classification system, method and device, a medium and a product, and relates to the field of cardiovascular disease identification, and the system comprises a data preprocessing module which carries out the filtering processing of an ECG signal, and reconstructs a pure signal; the feature extraction module extracts multi-scale features of the pure signals, determines heart rate variability features of a time domain, a frequency domain and a nonlinear domain, and generates multi-dimensional physiological feature vectors; the model learning optimization module is used for guiding a deep neural network model to learn discriminant feature expression with distinguishing ability by contrastively learning multi-dimensional physiological feature vectors in a model learning stage, and establishing a space class center based on a class center limiting mechanism in a model optimization stage; and the confidence interval decision module calculates the minimum class center distance from the ECG signal to be detected to all space class centers, constructs a confidence interval, classifies the ECG signal to be detected, and determines the type of the ECG signal. According to the invention, the ECG signal type which is not seen or unknown can be identified, and the misjudgment probability is reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of cardiovascular disease recognition, and in particular to an ECG signal classification system, method, device, medium and product. BACKGROUND

[0002] Cardiovascular disease (CVD) is a major public health problem worldwide. Early diagnosis and timely warning are of great significance to reduce the mortality rate of cardiovascular disease. Since cardiovascular disease can cause myocardial damage or cause arrhythmia, thereby changing the electrical activity of the heart, the morphological changes (i.e. waveform morphology) of electrocardiogram (ECG) signals can be used to identify certain diseases, such as atrial fibrillation. Existing recognition technologies, such as automatic electrocardiogram classification models, provide valuable auxiliary information for assisting disease diagnosis.

[0003] Classification models rely on labeling each known ECG signal for annotation, and then fitting. Although it has good recognition performance, but the ECG dataset with good annotation usually only covers limited ECG signal types. The traditional classification model trained with limited ECG signal types can only identify those ECG signal types that have been observed in the training set, and cannot identify unseen or unknown ECG signal types not included in the training data. But due to the spatial closure of the traditional classification model, if the patient has a cardiovascular disease type other than the model training, the trained classifier cannot identify it, and will incorrectly identify it as one of its known classes, resulting in misdiagnosis of the patient's disease, which seriously reduces the reliability of the model in clinical scenarios.

[0004] Therefore, in clinical applications, a method is needed that can identify unseen or unknown ECG signal types on the basis of known disease recognition of cardiovascular disease patients, so as to not only classify known classes, but also attribute all unseen or unknown data to a non-visible class, providing more valuable information for doctors. SUMMARY

[0005] The purpose of the present application is to provide an ECG signal classification system, method, device, medium and product to solve the problem of misdiagnosis caused by the inability to identify unseen or unknown ECG signal types.

[0006] To achieve the above purpose, the present application provides the following solutions.

[0007] In a first aspect, the present application provides an ECG signal classification system, comprising the following modules.

[0008] A data preprocessing module for filtering and reconstructing the pure signal of the ECG signal.

[0009] a feature extraction module configured to perform multi-scale feature extraction on the pure signal, determine heart rate variability features in time domain, frequency domain and nonlinear domain, and generate a multi-dimensional physiological feature vector suitable for input of a deep neural network model according to the heart rate variability features.

[0010] a model learning optimization module configured to, in a model learning stage, guide the deep neural network model to learn discriminative feature expression with distinguishing ability by comparing the multi-dimensional physiological feature vector, and in a model optimization stage, establish a spatial class center based on a class center restriction mechanism and optimize the deep neural network model; the multi-dimensional physiological feature vector includes positive samples and negative samples.

[0011] a confidence interval decision module configured to calculate minimum class center distances of a to-be-detected ECG signal to all spatial class centers, construct a confidence interval, and classify the to-be-detected ECG signal according to the minimum class center distances and the confidence interval to determine an ECG signal type; the ECG signal type includes a known class and an unknown class.

[0012] In a second aspect, the present application provides an ECG signal classification method, including the following steps.

[0013] based on a data preprocessing module, filtering and processing the ECG signal to reconstruct a pure signal.

[0014] based on a feature extraction module, performing multi-scale feature extraction on the pure signal, determining heart rate variability features in time domain, frequency domain and nonlinear domain, and generating a multi-dimensional physiological feature vector suitable for input of a deep neural network model according to the heart rate variability features.

[0015] based on a model learning optimization module, in a model learning stage, guiding the deep neural network model to learn discriminative feature expression with distinguishing ability by comparing the multi-dimensional physiological feature vector, and in a model optimization stage, establishing a spatial class center based on a class center restriction mechanism and optimizing the deep neural network model; the multi-dimensional physiological feature vector includes positive samples and negative samples.

[0016] based on a confidence interval decision module, calculating minimum class center distances of a to-be-detected ECG signal to all spatial class centers, constructing a confidence interval, and classifying the to-be-detected ECG signal according to the minimum class center distances and the confidence interval to determine an ECG signal type; the ECG signal type includes a known class and an unknown class.

[0017] In a third aspect, the present application provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the ECG signal classification method.

[0018] In a fourth aspect, the present application provides a computer readable storage medium, having stored thereon a computer program, which, when executed by a processor, implements the ECG signal classification method.

[0019] In a fifth aspect, the present application provides a computer program product, comprising a computer program, which, when executed by a processor, implements the ECG signal classification method.

[0020] According to the specific embodiments provided by the present application, the present application has the following technical effects: compared with the technical solution of recognizing diseases based on basic morphological changes, the present application comprehensively captures the multi-domain HRV features of the ECG signal, improves the feature representation capability, and at the same time, based on the contrast learning and the class center restriction mechanism, through the deep neural network model and the confidence interval, solves the problem that the traditional model can only identify the known types of ECG signals and misjudges the unknown types, significantly improves the clinical reliability, thereby identifies the unseen or unknown ECG signal types, and reduces the misjudgment probability. BRIEF DESCRIPTION OF DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of these drawings.

[0022] Figure 1 A structural schematic diagram of an ECG signal classification system provided by an embodiment of the present application.

[0023] Figure 2 A flowchart of an ECG signal classification method provided by an embodiment of the present application.

[0024] Figure 3 A confusion matrix diagram of arrhythmia recognition provided by an embodiment of the present application.

[0025] Figure 4 A bar chart of recognition accuracy of each type of arrhythmia provided by an embodiment of the present application.

[0026] Figure 5The arrhythmia spatial clustering scatter plot based on confidence interval decision provided in an embodiment of the present application. DETAILED DESCRIPTION

[0027] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0028] In order to make the purposes, characteristics and advantages of the present application more obvious and easy to understand, the present application will be further described in detail below with reference to the drawings and specific embodiments.

[0029] As shown in the figure, Figure 1 The ECG signal classification system provided in an embodiment of the present application includes the following modules.

[0030] The data preprocessing module is configured to perform filtering processing on the ECG signal and reconstruct a pure signal.

[0031] The feature extraction module is configured to perform multi-scale feature extraction on the pure signal, determine heart rate variability features in time domain, frequency domain and nonlinear domain, and generate a multi-dimensional physiological feature vector suitable for input of a deep neural network model according to the heart rate variability features.

[0032] The model learning and optimization module is configured to, in a model learning stage, guide the deep neural network model to learn discriminative feature expression with distinguishing ability by comparing the multi-dimensional physiological feature vectors, and in a model optimization stage, establish a spatial class center based on a class center restriction mechanism and optimize the deep neural network model; the multi-dimensional physiological feature vector includes positive samples and negative samples.

[0033] The confidence interval decision module is configured to calculate minimum class center distances of a to-be-tested ECG signal to all spatial class centers, construct a confidence interval, and classify the to-be-tested ECG signal according to the minimum class center distances and the confidence interval to determine an ECG signal type; the ECG signal type includes a known class and an unknown class.

[0034] In an exemplary embodiment, the data preprocessing module specifically includes the following units.

[0035] The wavelet decomposition unit is configured to perform multi-layer wavelet decomposition on the ECG signal by using a wavelet filter, and decompose the ECG signal into approximation coefficients and detail coefficients; the approximation coefficients reflect low-frequency trends of the ECG signal; and the detail coefficients reflect high-frequency changes of the ECG signal.

[0036] The reconstruction unit is used to iteratively reconstruct the pure signal from the highest level to the lowest level based on the approximation coefficients and the detail coefficients.

[0037] In practical applications, the data preprocessing module uses wavelet filters to decompose the ECG signal.

[0038] At each decomposition level, the ECG signal is decomposed into low-frequency approximation coefficients and high-frequency detail coefficients. The approximation coefficients reflect the low-frequency trend portion of the signal, while the detail coefficients reflect the high-frequency variations.

[0039] The approximate and detail coefficients obtained from the decomposition are processed, such as through threshold quantization, to remove noise. By setting an appropriate threshold, the main features of the ECG signal can be preserved while suppressing noise. The processed coefficients are then used for inverse transform to reconstruct the clean signal.

[0040] The main steps include the following.

[0041] First, regarding the input ECG signal Performing j-level wavelet decomposition, the approximation coefficients of the j-th level and detail coefficient as follows.

[0042]

[0043]

[0044] Among them, the approximation coefficient Represents the low-frequency components of the ECG signal, preserving the main signal shape; detail coefficients High-frequency components representing the ECG signal, including noise and details; Let n be the low-pass filter corresponding to the scaling function used in the j-th level wavelet decomposition, used to extract the approximate (low-frequency) components of the signal, where n is uniformly defined as the index of the discrete time. is the high-pass filter corresponding to the wavelet function used in the j-th level wavelet decomposition, used to extract the detailed (high-frequency) components of the signal; j is the wavelet decomposition level index.

[0045] Next comes signal reconstruction:

[0046] in, The approximation coefficients of the (j-1)th layer are obtained from the reconstruction of the jth layer. In the inverse wavelet reconstruction process, the lower-level signals are gradually reconstructed by combining the approximation coefficients and detail coefficients of the jth layer. k is the discrete-time index or the position index of the filter coefficients, which is used to represent the position of the filter or signal in the convolution and reconstruction process.

[0047] iteratively reconstructs from the highest layer j to the bottom layer 0, and finally obtains the pure signal is the pure signal of the 0th layer, i.e., the denoised ECG signal, is the final output signal obtained by reverse reconstruction from the highest layer.

[0048] In an exemplary embodiment, the feature extraction module specifically comprises the following units.

[0049] The positioning unit is configured to position the R-wave peak of the pure signal by using a peak detection algorithm.

[0050] The multi-scale extraction unit is configured to extract, based on the R-wave peak, the heart rate variability features in the time domain, the frequency domain and the nonlinear domain of the pure signal according to the sequence of the original waveform and the RR interval of the pure signal.

[0051] The processing unit is configured to normalize or standardize the heart rate variability features to determine the processed heart rate variability features; wherein the RR interval is the time interval between adjacent R-wave peaks on an electrocardiogram.

[0052] The multi-dimensional physiological feature vector generation unit is configured to generate a multi-dimensional physiological feature vector suitable for the input of a deep neural network model according to the processed heart rate variability features.

[0053] In actual application, the feature extraction module first positions the R-wave peak by using the Pan-Tompkins peak detection algorithm on the pure signal; then extracts the HRV features in the time domain, the frequency domain and the nonlinear domain based on the sequence of the RR interval and the original waveform; and finally normalizes or standardizes the obtained HRV feature vector to generate a D-dimensional physiological feature vector suitable for the input of a deep neural network model.

[0054] The R-peak detection is performed on the preprocessed electrocardiogram waveform, and a double-threshold QRS wave detection algorithm with Pan-Tompkins adaptability is used to position the R-wave peak of the ECG signal, and the subsequent HRV feature extraction can be performed after accurately identifying the R-peak position.

[0055] The HRV time domain features (MeanNN, SDNN, RMSSD, NN50, pNN50), the frequency domain features (very low frequency VLF, low frequency power LF, high frequency power HF, LF / HF) and the nonlinear features (ApEn, SampEn, Poincaré SD1 / SD2) are extracted from the sequence of the RR interval and the original waveform to construct a D-dimensional physiological feature vector X d ∈R D The D-dimensional physiological feature vector X d ​is the HRV feature vector, R D is a D-dimensional real number space, representing the extracted D-dimensional physiological feature vector X d is a D-dimensional real number vector, belonging to the mathematical Euclidean space, i.e. D-dimensional real number space R D .

[0056] In an exemplary embodiment, the model learning optimization module specifically comprises the following units.

[0057] The standardization processing unit is configured to perform standardization processing on the multi-dimensional physiological feature vector to determine a processed feature vector.

[0058] The contrast learning unit is configured to perform contrast learning on the processed feature vector in the model learning stage to construct a contrast relationship between positive samples and negative samples; the positive samples are arrhythmia HRV features of the same type, and the arrhythmia HRV features of the same type include low HF features and high LF / HF features of atrial fibrillation patients; the negative samples are difficult samples with similar features but different diagnostic conclusions from the processed feature vector.

[0059] The guiding unit is configured to guide the deep neural network model to learn discriminative feature expression with distinguishing ability according to the contrast relationship, and determine a learned deep neural network model.

[0060] The space class center determination unit is configured to introduce a class center restriction mechanism in the model optimization stage, taking the mean vector of the arrhythmia HRV features of each type as a space class center.

[0061] The core loss function determination unit is configured to calculate the Euclidean distance between each sample and the corresponding space class center, and minimize the Euclidean distance between all samples and the corresponding space class center to compress the intra-class distribution, and determine a core loss function in the model optimization stage; the sample is the processed feature vector.

[0062] The optimization unit is configured to optimize the deep neural network model according to the core loss function.

[0063] In actual application, the D-dimensional physiological feature vector X d is standardized to realize equal proportion scaling of the original data.

[0064]

[0065] wherein, is the processed feature vector, is the s-th feature value in the original feature vector is the mean, is the standard deviation, and s is the index of the original feature vector.

[0066] By using the mean and standard deviation of variable values to convert the original data into data within a certain range, the dimension and order of magnitude are eliminated, and the weight of the variable in the analysis is changed to solve the problem of different measurements. Make it suitable for input into a deep neural network model (referred to as model), improve the training efficiency and stability of the model.

[0067] The model learning optimization module is divided into two stages of model learning and optimization. In the model learning stage, contrastive learning is performed using HRV features. The main body of learning is positive samples and negative samples. The positive samples are HRV features of the same arrhythmia (such as low HF features and high LF / HF features of atrial fibrillation patients). The negative samples are samples similar to the original sample in features but different in diagnosis conclusion, i.e. difficult samples. By constructing such a contrast relationship, the model is guided to learn discriminative feature expression.

[0068] In the model optimization stage, in order to make the HRV features of the same class more closely gathered in the model and more distant between different classes, a class center restriction mechanism is introduced, that is, the mean vector of the HRV features of each arrhythmia is taken as the class center. By minimizing the Euclidean distance between the sample embedding and the corresponding class center, the model compresses the intra-class feature distribution while maintaining the inter-class discrimination, enhancing the generalization ability of the model in identifying unknown type samples.

[0069] Among them, the construction of difficult samples first locates the diagnostic key features, and the diagnostic key features of the D-dimensional physiological feature vector X d are defined as , where t is the diagnostic key feature index. For example, the diagnostic key features of atrial fibrillation are lower HF features and higher LF / HF features compared to normal electrocardiogram. .

[0070] Then, a large perturbation is applied to the diagnostic key features. For each diagnostic key feature, is transformed according to the category in the opposite or enhanced trend.

[0071]

[0072] , where is the transformed original value, and β is the scaling factor controlling the perturbation amplitude, typically 3.

[0073] The overall framework of model learning and optimization first trains an initialization model based on a deep neural network model, which is composed of multiple layers of bidirectional long short-term memory network (Bi-LSTM) and one-dimensional convolutional neural network (1D-CNN). Among them, the number of hidden units of each layer of Bi-LSTM is 128, and the number of channels of 1D-CNN is 64, and the kernel width is 3.

[0074] Then the loss function of positive and negative sample pair comparison learning is constructed ,

[0075] Wherein, J={1,2,3,...,B} is the set of batch index, B is the total number of batch index, ={ ,y m =y c} represents the set of positive samples, m is the index number, y m represents the type, y c is the class label of the cth sample, y m =y c , that is, the same type; { ,y m ≠y c} represents the set of negative samples, q is the sample index in the negative sample set; is the multi-dimensional physiological feature vector of the mth sample, ∈R D ; is the feature vector of the pth positive sample; is the feature vector of the bth sample (positive or negative sample); b is the index of positive or negative sample, and all positive and negative samples are traversed; p is the index of positive sample, and all positive sample samples are traversed; is the vector inner product calculation, which is used to measure the similarity between representations; is the temperature parameter to control the sharpness of the similarity distribution.

[0076] Then the model optimization stage is introduced, and the class center restriction mechanism is introduced, taking the cth class as an example, let all the training samples of the cth class be mapped to{ } after mapping, define as the spatial restriction center point of the class, that is, the spatial class center of the cth class, and the average of the representation vectors extracted from the HRV features of the arrhythmia samples in the class in the model, wherein C is the total number of known categories, that is, the number of all known categories.

[0077]

[0078] Wherein, is the training data of the cth class.

[0079] Then the Euclidean distance between each sample and its corresponding class center is calculated, and the Euclidean distance is the straight line distance between two points in the feature space, that is The Euclidean distance between and is defined as follows.

[0080] =

[0081] where D is the dimension of the HRV feature vector, is the e-th feature value of the c-th sample. is the e-th coordinate value of the c-th class center, e is the feature dimension index.

[0082] By minimizing the Euclidean distance between all samples and their corresponding class centers, the intra-class distribution is compressed, and the core loss function of the model optimization stage is ,

[0083] where, is the regularization coefficient, is all the learnable network parameters of the feature extractor. In this way, by minimizing the objective function, the same class of samples can be concentrated around their class centers, enhancing the model's ability to distinguish different arrhythmias and improving the robustness of identifying unknown classes.

[0084] The model learning optimization module learns the center vector of each arrhythmia class at the end of the deep network through contrast learning (positive and negative sample contrast) and class center restriction mechanism, and minimizes the loss function between the same class samples and their class centers , thereby compressing the intra-class distribution and expanding the inter-class difference in the feature space.

[0085] In an exemplary embodiment, the deep neural network model is composed of a multi-layer bidirectional long short-term memory network and a one-dimensional convolutional neural network.

[0086] In an exemplary embodiment, the confidence interval decision module specifically includes the following modules.

[0087] The estimation unit is configured to calculate the minimum class center distance of the to-be-tested ECG signal to all spatial class centers and estimate the sample variance of the to-be-tested ECG signal.

[0088] The confidence interval construction unit is configured to construct a confidence interval according to the sample variance and the hyperparameters of the optimized deep neural network model.

[0089] The judgment unit is configured to judge whether the minimum class center distance exceeds the confidence interval.

[0090] The unknown class determination unit is configured to determine that the to-be-tested ECG signal is an unknown class when the output result of the judgment unit is yes.

[0091] The known class determination unit is configured to determine the ECG signal to be tested as a known class when the output result of the determination unit is no, and determine the accurate known type of the ECG signal to be tested by using statistical analysis information.

[0092] In practical applications, after the model learning and model optimization are completed, the confidence interval decision module is responsible for mapping the embedding representation of each test sample to the "known class" or "unknown class", which is specifically divided into the following three steps.

[0093] First, the class center mentioned above and the Euclidean distance are used, that is, the reference for calculating the Euclidean distance of each sample in the class to the center .

[0094]

[0095] wherein, is the Euclidean distance from a sample in the c-th class to the spatial class center of the class.

[0096] Estimate the sample variance as follows.

[0097]

[0098] wherein, is the average value of the Euclidean distance from a sample in the c-th class to the spatial class center of the class.

[0099] The distance of the test sample to the decision, for any test sample x, the embedding representation z is extracted by the model. Calculate the minimum distance of the embedding to all known class centers .

[0100]

[0101] wherein, C is the total number of known classes, that is, the number of all known classes.

[0102] Finally, the decision criterion for setting the confidence interval is set, and the hyperparameter is set, generally , which corresponds to the multiple threshold value of the approximate 95% confidence interval.

[0103] If the distance of the test sample to the nearest class center exceeds the confidence interval of the distance distribution of the class.

[0104] That is, .

[0105] wherein, ​​For the confidence interval, the confidence interval is used in the system to measure whether the distance of a test sample from its nearest class center belongs to the normal (training) distance distribution range of the class. If it exceeds, it is judged as unknown class; The class index with the minimum distance of the test sample (the class number of the nearest class center).

[0106] Then it is determined as an unknown class; otherwise, it is classified into the nearest known class. The overall analysis logic is as follows.

[0107]

[0108] The confidence interval decision module can not only ensure accurate classification of known class samples by using statistical distribution information, but also safely identify them as "unknown" type when the samples significantly deviate from any class center, i.e., unknown class, to provide reliable prompt for subsequent clinical review.

[0109] The confidence interval decision module is based on the loss function of each class sample in the training set to the center Distance distribution, calculate the mean and variance and set a multiple sigma threshold to form the confidence interval. If the test sample is more than the threshold distance from its nearest center, it is judged as unknown class, otherwise it is classified into the corresponding known class, realizing the smooth connection from closed set classification to open set detection.

[0110] As shown in Figure 2 The present application provides an ECG signal classification method, comprising the following steps.

[0111] S1: Based on the data preprocessing module, filter the ECG signal and reconstruct the pure signal.

[0112] S2: Based on the feature extraction module, multi-scale feature extraction is performed on the pure signal to determine the heart rate variability features in time domain, frequency domain and nonlinear domain, and a multi-dimensional physiological feature vector suitable for input of a deep neural network model is generated according to the heart rate variability features.

[0113] S3: Based on the model learning optimization module, in the model learning stage, the multi-dimensional physiological feature vector is compared to guide the deep neural network model to learn discriminative feature expression with distinguishing ability, and in the model optimization stage, based on the class center restriction mechanism, the space class center is established to optimize the deep neural network model; the multi-dimensional physiological feature vector includes positive samples and negative samples.

[0114] S4: Based on the confidence interval decision module, calculate the minimum class center distance of the ECG signal to be tested to all space class centers, construct the confidence interval, and classify the ECG signal to be tested according to the minimum class center distance and the confidence interval to determine the ECG signal type; the ECG signal type includes known class and unknown class.

[0115] The prior art mainly focuses on the optimization of electrocardiosignal acquisition hardware (such as multi-channel synchronization, wearable design), basic waveform matching (such as pattern comparison, threshold judgment) or single disease classification (such as atrial fibrillation recognition), and has the following common limitations.

[0116] (1) Shallow feature extraction: relying on waveform morphology or simple time-frequency features, without integrating multi-domain HRV features (time domain, frequency domain, nonlinearity).

[0117] (2) Closed classification: only able to identify known types in the training set, prone to misjudgment for unknown arrhythmias.

[0118] (3) Static decision-making: lack of confidence interval mechanism, unable to quantify classification uncertainty.

[0119] Based on the technical solution of the present application, the present application can achieve the following effects in view of the above technical problems.

[0120] (1) Confidence interval decision-making module.

[0121] Effect: Solves the misjudgment problem of traditional models for unknown arrhythmia types, significantly improves clinical reliability.

[0122] Data support: The identification accuracy rate for unknown categories reaches 95%, as shown in the confusion matrix, avoiding incorrect classification as known types. Figure 3

[0123] By estimating the confidence interval through statistical distribution, the classification error rate of known categories is less than 5%, as shown in the category accuracy rate. Figure 4

[0124] (2) Model learning optimization module: contrastive learning and class center restriction mechanism.

[0125] Effect: Enhances feature discrimination, compresses intra-class distribution and expands inter-class distance.

[0126] Data support: Known category classification accuracy rate improved: normal rhythm (Normal) 92%, atrial fibrillation (AF) 93%, as shown. Figure 4

[0127] The feature space visually displays the tight clustering of intra-class samples, Figure 5 as shown, with clear inter-class boundaries.

[0128] (3) Feature extraction module: multi-dimensional HRV feature extraction.

[0129] Effect: Comprehensive capture of time domain, frequency domain and nonlinear characteristics of electrocardiosignal, improving feature representation ability.

[0130] ​​​Data support: combined with 128-dimensional Bi-LSTM and 64-channel 1D-CNN, the feature discrimination is improved by about 12% compared with a single model (comparative experiment not shown).

[0131] The present application covers time domain, frequency domain, nonlinear and morphological multi-angle features, adopts a deep neural network model of Bi-LSTM+1D-CNN, and realizes unknown class detection through confidence interval expansion, so as to realize abnormal identification of cardiovascular diseases from known to unknown by the model. This method not only improves the early diagnosis rate of cardiovascular diseases, but also can efficiently process and analyze huge electrocardiogram data in the background of big data era, provide personalized health management services, and ultimately promote the prevention and treatment of cardiovascular diseases, so as to realize early diagnosis and personalized health management. This method provides new possibilities for the prevention and treatment of cardiovascular diseases in portability, intelligence and big data processing capacity.

[0132] In an exemplary embodiment, a computer device is provided, comprising a memory and a processor, the memory storing a computer program, and the processor implementing the steps in the above method embodiments when executing the computer program. The computer device can be a server or a terminal. The computer device comprises a processor, a memory, an input / output interface (I / O) and a communication interface. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device comprises a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operating system and the computer program in the non-volatile storage medium to run. The database of the computer device is used to store data to be processed. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with external terminals through network connection. The computer program is executed by the processor to implement an ECG signal classification method.

[0133] In an exemplary embodiment, a computer device is provided, comprising a memory and a processor, the memory storing a computer program, and the processor implementing the steps in the above method embodiments when executing the computer program.

[0134] In an exemplary embodiment, a computer readable storage medium is provided, storing a computer program, and the computer program is executed by a processor to implement the steps in the above method embodiments.

[0135] In an example embodiment, a computer program product is provided, including a computer program which, when executed by a processor, implements the steps of any of the above method embodiments.

[0136] It should be noted that the user information (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or authorized by all parties, and the collection, use and processing of related data need to comply with relevant regulations.

[0137] It can be understood by those skilled in the art that all or part of the processes in the above-mentioned embodiments can be completed by computer program instruction related hardware. The computer program can be stored in a non-volatile computer readable storage medium. When the computer program is executed, it can include the processes of the above-mentioned embodiments. Any reference to memory, database or other medium used in the embodiments provided by the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical storage, high-density embedded non-volatile memory, resistive memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration but not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0138] The database involved in the embodiments provided by the present application can include at least one of a relational database and a non-relational database. The non-relational database can include a distributed database based on blockchain, etc., without being limited thereto. The processor involved in the embodiments provided by the present application can be a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., without being limited thereto.

[0139] Any technical features in the above embodiments can be combined, and for the sake of brevity, not all possible combinations are described above, however, it should be understood that the application encompasses all possible combinations of the technical features described above.

[0140] The principles and implementation manners of the present application are described herein by using specific examples, and the above embodiments are only used to help understand the method of the present application and its core idea; meanwhile, according to the idea of the present application, the specific implementation manners and application scopes will be changed by those skilled in the art. In conclusion, the content of the present specification should not be understood as a limitation of the present application.

Claims

1. An ECG signal classification system, characterized in that, include: The data preprocessing module is used to filter the ECG signal and reconstruct a clean signal; The feature extraction module is used to perform multi-scale feature extraction on the pure signal, determine the heart rate variability features in the time domain, frequency domain and nonlinear domain, and generate a multi-dimensional physiological feature vector suitable for input to a deep neural network model based on the heart rate variability features. The model learning and optimization module is used to guide the deep neural network model to learn discriminative feature representations with distinguishing ability by comparing and learning the multidimensional physiological feature vectors during the model learning stage, and to establish spatial class centers and optimize the deep neural network model based on the class center constraint mechanism during the model optimization stage; the multidimensional physiological feature vectors include positive samples and negative samples; The confidence interval decision module is used to calculate the minimum class center distance from the ECG signal under test to all spatial class centers, construct a confidence interval, and classify the ECG signal under test according to the minimum class center distance and the confidence interval to determine the ECG signal type; the ECG signal type includes known class and unknown class.

2. The ECG signal classification system according to claim 1, characterized in that, The data preprocessing module specifically includes: The wavelet decomposition unit is used to perform multi-level wavelet decomposition on the ECG signal using a wavelet filter, decomposing the ECG signal into approximation coefficients and detail coefficients; the approximation coefficients reflect the low-frequency trend of the ECG signal; and the detail coefficients reflect the high-frequency changes of the ECG signal. The reconstruction unit is used to iteratively reconstruct the pure signal from the highest level to the lowest level based on the approximation coefficients and the detail coefficients.

3. The ECG signal classification system according to claim 1, characterized in that, The feature extraction module specifically includes: A positioning unit is used to locate the R-wave peak of the pure signal using a peak detection algorithm; A multi-scale extraction unit is used to extract the heart rate variability features of the pure signal in the time domain, frequency domain, and nonlinear domain based on the R-wave peak and according to the original waveform and RR interval sequence of the pure signal. The processing unit is used to normalize or standardize the heart rate variability characteristics to determine the processed heart rate variability characteristics; wherein, the RR interval is the time interval between adjacent R wave peaks on the electrocardiogram; A multidimensional physiological feature vector generation unit is used to generate a multidimensional physiological feature vector suitable for input to a deep neural network model based on the processed heart rate variability features.

4. The ECG signal classification system according to claim 1, characterized in that, The model learning and optimization module specifically includes: A standardization processing unit is used to standardize the multidimensional physiological feature vectors and determine the processed feature vectors. The contrastive learning unit is used to perform contrastive learning on the processed feature vector during the model learning phase to construct a contrastive relationship between positive and negative samples. The positive samples are heart rate variability features of the same type of arrhythmia, including low HF features and high LF / HF features of atrial fibrillation patients. The negative samples are difficult samples that have similar features to the processed feature vector but have different diagnostic conclusions. The guiding unit is used to guide the deep neural network model to learn discriminative feature representations with distinguishing ability based on the comparison relationship, and to determine the learned deep neural network model. The spatial class center determination unit is used to introduce a class center constraint mechanism during the model optimization stage, using the mean vector of heart rate variability features of various arrhythmias as the spatial class center; The core loss function determination unit is used to calculate the Euclidean distance between each sample and the corresponding spatial class center, minimize the Euclidean distance between all samples and their corresponding centers, compress the intra-class distribution, and determine the core loss function for the model optimization stage; the sample is the processed feature vector. An optimization unit is used to optimize the deep neural network model based on the core loss function.

5. The ECG signal classification system according to claim 4, characterized in that, The deep neural network model is composed of a multi-layer bidirectional long short-term memory network and a one-dimensional convolutional neural network.

6. The ECG signal classification system according to claim 1, characterized in that, The confidence interval decision module specifically includes: The estimation unit is used to calculate the minimum class center distance from the ECG signal under test to all spatial class centers, and to estimate the sample variance of the ECG signal under test. The confidence interval construction unit is used to construct confidence intervals based on the sample variance and the hyperparameters of the optimized deep neural network model. The judgment unit is used to determine whether the minimum class center distance exceeds the confidence interval; An unknown class determination unit is used to determine that the ECG signal to be tested is of an unknown class when the output result of the determination unit is yes. A known class determination unit is used to determine that the ECG signal under test is of a known class when the output result of the judgment unit is negative, and to determine the precise known type of the ECG signal under test using statistical analysis information.

7. An ECG signal classification method, characterized in that, The ECG signal classification method is applied to the ECG signal classification system according to any one of claims 1-6, and the ECG signal classification method includes: Based on the data preprocessing module, the ECG signal is filtered to reconstruct a clean signal; Based on the feature extraction module, multi-scale feature extraction is performed on the pure signal to determine the heart rate variability features in the time domain, frequency domain, and nonlinear domain, and a multi-dimensional physiological feature vector suitable for input to a deep neural network model is generated based on the heart rate variability features. Based on the model learning and optimization module, during the model learning phase, the deep neural network model is guided to learn discriminative feature representations with distinguishing capabilities by comparing and learning the multidimensional physiological feature vectors. During the model optimization phase, spatial class centers are established based on the class center constraint mechanism to optimize the deep neural network model. The multidimensional physiological feature vectors include positive samples and negative samples. Based on the confidence interval decision module, the minimum class center distance from the ECG signal under test to all spatial class centers is calculated, a confidence interval is constructed, and the ECG signal under test is classified according to the minimum class center distance and the confidence interval to determine the ECG signal type; the ECG signal type includes known class and unknown class.

8. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the ECG signal classification method of claim 7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the ECG signal classification method of claim 7.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the ECG signal classification method of claim 7.

Images