Electrocardiogram data processing method and device and storage medium

Abstract

The invention discloses an electrocardiogram data processing method and device and a storage medium, and belongs to the technical field of artificial intelligence. The target electrocardio sequence is input into an electrocardio data processing model for electrocardio information mining, cardiac beat distribution information and event information of cardiac electrophysiological activities are obtained, the cardiac beat distribution information comprises a cardiac beat type in the target electrocardio sequence, and the event information comprises an event coverage area and an event type; wherein the electrocardiogram data processing model is an end-to-end model for synchronously outputting the heart beat type, the event coverage area and the event type. According to the embodiment of the invention, the end-to-end electrocardio data processing model is obtained through direct training without depending on other electrocardio signal analysis methods, and the heart beat distribution information and the heart electrophysiological activity event information in the input target electrocardio sequence can be directly output in an end-to-end manner for the input target electrocardio sequence; and the mining accuracy of the event information of the heart electrophysiological activity is greatly improved.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to an electrocardiogram data processing method, device and storage medium. Background Art

[0002] ECG data is an important type of medical data, but related technologies are not capable of information mining from ECG data. For example, related technologies usually can only rely on artificial intelligence models that do not have a global perspective to identify a single heartbeat in an ECG data segment of fixed length, and then combine the identified single heartbeat with other ECG data analysis methods to further analyze more event information of cardiac electrophysiological activities. It is impossible to directly mine a variety of information including event information of cardiac electrophysiological activities based on an end-to-end artificial intelligence model. Moreover, the fixed-length ECG data segment is short, usually only about 30 seconds, and the effective information content for analyzing event information of cardiac electrophysiological activities is insufficient, which reduces the recognition accuracy of event information of cardiac electrophysiological activities.

[0003] Furthermore, when analyzing the event information of cardiac electrophysiological activity, the related technologies usually need to rely on assumed information, for example, assuming that the position of the heartbeat has a certain regularity, and then the event information of cardiac electrophysiological activity can be analyzed. However, when some abnormal events of cardiac electrophysiological activity occur, this assumption does not hold, which directly leads to inaccurate mining results of the related technologies for the event information of cardiac electrophysiological activity. Summary of the invention

[0004] The embodiments of the present application provide an electrocardiogram data processing method, device, and storage medium to solve at least one of the aforementioned technical problems.

[0005] According to one aspect of an embodiment of the present application, a method for processing electrocardiogram data is provided, the method comprising:

[0006] Acquire target ECG sequence;

[0007] Inputting the target ECG sequence into an ECG data processing model to perform ECG information mining to obtain heartbeat distribution information and event information of cardiac electrophysiological activities, wherein the heartbeat distribution information includes the heartbeat type in the target ECG sequence, and the event information includes the event coverage area and the event type;

[0008] The electrocardiogram data processing model is an end-to-end model that synchronously outputs the heartbeat type, the event coverage area, and the event type.

[0009] In an exemplary embodiment, the method further comprises:

[0010] Based on the event information, calibrating the heartbeat distribution information to obtain calibrated heartbeat distribution information;

[0011] or,

[0012] Based on the heartbeat distribution information, the event information is calibrated to obtain calibrated event information.

[0013] In an exemplary embodiment, the ECG data processing model is trained by the following method:

[0014] Acquire a sample ECG sequence, and acquire heart beat reference information and event reference information corresponding to the sample ECG sequence;

[0015] Constructing a preset model, wherein the preset model includes a global ECG feature extraction module, a local ECG feature extraction module, an ECG feature fusion module, an event location module, an event classification module, and a heart beat classification module;

[0016] Based on the global ECG feature extraction module, extract global ECG features from the sample ECG sequence to obtain global ECG features;

[0017] Based on the local ECG feature extraction module, extracting local ECG features from the global ECG features to obtain local ECG features;

[0018] Inputting the global ECG feature and the local ECG feature into the ECG feature fusion module to perform ECG feature fusion to obtain a fused ECG feature;

[0019] Inputting the fused ECG features into the heartbeat classification module, the event location module and the event classification module respectively, performing heartbeat classification, event location and event classification respectively, and obtaining sample heartbeat distribution information and sample event information of cardiac electrophysiological activities;

[0020] determining a first loss based on a difference between the sample heartbeat distribution information and the heartbeat reference information;

[0021] determining a second loss based on a difference between the sample event information and the event reference information;

[0022] Based on the first loss and the second loss, the parameters of each module in the preset model are adjusted to obtain the electrocardiogram data processing model.

[0023] In an exemplary embodiment, the method further comprises:

[0024] Dividing the sample ECG sequence into a plurality of sample ECG segments of equal length based on a first preset step length;

[0025] Dividing each of the sample ECG segments into a plurality of sample ECG regions of equal length based on a second preset step length;

[0026] The heartbeat reference information includes a reference heartbeat type corresponding to each of the sample ECG regions; and the event reference information includes a reference event coverage area and a reference event type corresponding to each of the sample ECG segments.

[0027] In an exemplary embodiment, the method further comprises:

[0028] The first preset step size and the second preset step size are determined based on the change in the time scale during the global electrocardiogram feature extraction process.

[0029] In an exemplary embodiment, the sample heartbeat distribution information includes a predicted heartbeat type corresponding to each of the sample electrocardiogram regions, and determining the first loss based on a difference between the sample heartbeat distribution information and the heartbeat reference information includes:

[0030] For any of the sample ECG regions, determining a corresponding heartbeat type prediction loss based on a difference between a reference heartbeat type corresponding to the sample ECG region and a corresponding predicted heartbeat type;

[0031] The first loss is determined based on the predicted loss of the heartbeat types corresponding to the sample electrocardiogram regions.

[0032] In an exemplary embodiment, the sample event information includes a predicted event coverage area and a predicted event type corresponding to each of the sample ECG segments, and determining the second loss based on the difference between the sample event information and the event reference information includes:

[0033] For any of the sample ECG segments, based on the difference between the reference event coverage area corresponding to the sample ECG segment and the corresponding predicted event coverage area, determine the corresponding event positioning loss; and based on the difference between the reference event type corresponding to the sample ECG segment and the corresponding predicted event type, determine the corresponding event type prediction loss;

[0034] The second loss is determined based on the event localization losses respectively corresponding to the sample ECG segments and the event type prediction losses respectively corresponding to the sample ECG segments.

[0035] In an exemplary embodiment, the global ECG feature extraction module includes at least two convolutional layers for generating a time-scale variation;

[0036] The local ECG feature extraction module includes at least two bidirectional long short-term memory network layers;

[0037] The ECG feature fusion module is used to splice the features output by the last network layer of the global ECG feature extraction module with the features output by the last network layer of the local ECG feature extraction module, and perform feature fusion on the splicing result.

[0038] According to one aspect of an embodiment of the present application, there is provided an electrocardiogram data processing device, the device comprising:

[0039] A target ECG sequence acquisition module is used to acquire a target ECG sequence;

[0040] A target ECG sequence processing module is used to input the target ECG sequence into an ECG data processing model to perform ECG information mining to obtain heartbeat distribution information and event information of cardiac electrophysiological activities, wherein the heartbeat distribution information includes the heartbeat type in the target ECG sequence, and the event information includes event coverage area and event type;

[0041] The electrocardiogram data processing model is an end-to-end model that synchronously outputs the heartbeat type, the event coverage area, and the event type.

[0042] In an exemplary embodiment, the target ECG sequence processing module is used to perform the following operations:

[0043] Based on the event information, calibrating the heartbeat distribution information to obtain calibrated heartbeat distribution information;

[0044] or,

[0045] Based on the heartbeat distribution information, the event information is calibrated to obtain calibrated event information.

[0046] In an exemplary embodiment, the target ECG sequence processing module is used to perform the following operations:

[0047] Acquire a sample ECG sequence, and acquire heart beat reference information and event reference information corresponding to the sample ECG sequence;

[0048] Constructing a preset model, wherein the preset model includes a global ECG feature extraction module, a local ECG feature extraction module, an ECG feature fusion module, an event location module, an event classification module, and a heart beat classification module;

[0049] Based on the global ECG feature extraction module, extract global ECG features from the sample ECG sequence to obtain global ECG features;

[0050] Based on the local ECG feature extraction module, extracting local ECG features from the global ECG features to obtain local ECG features;

[0051] Inputting the global ECG feature and the local ECG feature into the ECG feature fusion module to perform ECG feature fusion to obtain a fused ECG feature;

[0052] Inputting the fused ECG features into the heartbeat classification module, the event location module and the event classification module respectively, performing heartbeat classification, event location and event classification respectively, and obtaining sample heartbeat distribution information and sample event information of cardiac electrophysiological activities;

[0053] determining a first loss based on a difference between the sample heartbeat distribution information and the heartbeat reference information;

[0054] determining a second loss based on a difference between the sample event information and the event reference information;

[0055] Based on the first loss and the second loss, the parameters of each module in the preset model are adjusted to obtain the electrocardiogram data processing model.

[0056] In an exemplary embodiment, the target ECG sequence processing module is used to perform the following operations:

[0057] Dividing the sample ECG sequence into a plurality of sample ECG segments of equal length based on a first preset step length;

[0058] Dividing each of the sample ECG segments into a plurality of sample ECG regions of equal length based on a second preset step length;

[0059] The heartbeat reference information includes a reference heartbeat type corresponding to each of the sample ECG regions; and the event reference information includes a reference event coverage area and a reference event type corresponding to each of the sample ECG segments.

[0060] In an exemplary embodiment, the target ECG sequence processing module is used to perform the following operations:

[0061] The first preset step size and the second preset step size are determined based on the change in the time scale during the global electrocardiogram feature extraction process.

[0062] In an exemplary embodiment, the sample heartbeat distribution information includes the predicted heartbeat type corresponding to each sample ECG region. The target ECG sequence processing module is used to perform the following operations:

[0063] For any of the sample ECG regions, determining a corresponding heartbeat type prediction loss based on a difference between a reference heartbeat type corresponding to the sample ECG region and a corresponding predicted heartbeat type;

[0064] The first loss is determined based on the predicted loss of the heartbeat types corresponding to the sample electrocardiogram regions.

[0065] In an exemplary embodiment, the sample event information includes the predicted event coverage area and predicted event type corresponding to each sample ECG segment, and the target ECG sequence processing module is used to perform the following operations:

[0066] For any of the sample ECG segments, based on the difference between the reference event coverage area corresponding to the sample ECG segment and the corresponding predicted event coverage area, determine the corresponding event positioning loss; and based on the difference between the reference event type corresponding to the sample ECG segment and the corresponding predicted event type, determine the corresponding event type prediction loss;

[0067] The second loss is determined based on the event localization losses respectively corresponding to the sample ECG segments and the event type prediction losses respectively corresponding to the sample ECG segments.

[0068] In an exemplary embodiment, the global ECG feature extraction module includes at least two convolutional layers for generating a time-scale variation;

[0069] The local ECG feature extraction module includes at least two bidirectional long short-term memory network layers;

[0070] The ECG feature fusion module is used to splice the features output by the last network layer of the global ECG feature extraction module with the features output by the last network layer of the local ECG feature extraction module, and perform feature fusion on the splicing result.

[0071] According to one aspect of an embodiment of the present application, a computer device is provided, comprising a processor and a memory, wherein the memory stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set are loaded and executed by the processor to implement the above-mentioned electrocardiogram data processing method.

[0072] According to one aspect of an embodiment of the present application, a computer-readable storage medium is provided, in which at least one instruction, at least one program, a code set or an instruction set is stored. The at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by a processor to implement the above-mentioned electrocardiogram data processing method.

[0073] According to one aspect of the embodiments of the present application, a computer program product is provided, the computer program product comprising computer instructions, the computer instructions being stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes to implement the above-mentioned electrocardiogram data processing method.

[0074] The technical solution provided in the embodiments of the present application can bring the following beneficial effects:

[0075] The embodiments of the present application propose an ECG data processing method, device, and storage medium. The embodiments of the present application directly train an end-to-end ECG data processing model without relying on other ECG signal analysis methods. For an input target ECG sequence, the beat distribution information and event information of the cardiac electrophysiological activity in the target ECG sequence can be directly output end-to-end. Moreover, the ECG data processing model has a global vision, has no restrictions on the length of the input target ECG sequence, and does not need to be cut into ECG data segments of fixed length, and can be directly input into the ECG data processing model. The ECG data processing model directly mines the event information of the cardiac electrophysiological activity in an end-to-end mode without relying on assumed information, greatly improving the accuracy of mining the event information of the cardiac electrophysiological activity. BRIEF DESCRIPTION OF THE DRAWINGS

[0076] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0077] Figure 1 is a schematic diagram of an application program operating environment provided by an embodiment of the present application;

[0078] Figure 2 is a flow chart of an electrocardiogram data processing method provided by an embodiment of the present application;

[0079] Figure 3 This is a flowchart of an electrocardiogram data processing model training method provided by an embodiment of the present application;

[0080] Figure 4 It is a schematic diagram of the architecture of a preset model provided by an embodiment of the present application;

[0081] Figure 5 It is a flowchart of a first loss determination method provided by an embodiment of the present application;

[0082] Figure 6 is a flow chart of a second loss determination method provided by an embodiment of the present application;

[0083] Figure 7 is a block diagram of an electrocardiogram data processing device provided by an embodiment of the present application;

[0084] Figure 8It is a structural block diagram of a computer device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0085] To make the purpose, technical solution and advantages of this application clearer, the following will further describe the implementation of this application in detail with reference to the accompanying drawings. It should be noted that the various data used in the embodiments of this application have been fully authorized by the relevant entities before use.

[0086] Please refer to Figure 1 , which shows a schematic diagram of an application program running environment provided by an embodiment of the present application under an exemplary implementation. The application program running environment may include: a terminal 10 and a server 20.

[0087] The terminal 10 includes, but is not limited to, electronic devices such as mobile phones, computers, intelligent voice interaction devices, smart home appliances, vehicle-mounted terminals, game consoles, e-book readers, multimedia playback devices, wearable devices, etc. The terminal 10 may be installed with a client of an application program.

[0088] In the embodiment of the present application, the above-mentioned application can be any application that runs an ECG data processing service and supports ECG data processing. Typically, the application is an intelligent medical application. Of course, in addition to intelligent medical applications, other types of applications can also run ECG data processing services and support ECG data processing. For example, virtual reality (VR) applications, augmented reality (AR) applications, etc., which are not limited in the embodiment of the present application. Optionally, a client of the above-mentioned application is running in the terminal 10.

[0089] The server 20 is used to provide background services for the client of the application in the terminal 10. For example, the server 20 can be the background server of the above application. The server 20 can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms. Optionally, the server 20 provides background services for applications in multiple terminals 10 at the same time.

[0090] Optionally, the terminal 10 and the server 20 may communicate with each other via a network 30. The terminal 10 and the server 20 may be directly or indirectly connected via wired or wireless communication, which is not limited in the present application.

[0091] Please refer to Figure 2 , which shows a flowchart of an exemplary embodiment of an ECG data processing method. The method can be applied to a computer device, which refers to an electronic device with data calculation and processing capabilities. For example, the execution subject of each step can be Figure 1 The method may include the following steps:

[0092] Step S201: Acquire a target ECG sequence.

[0093] The target ECG sequence may be a sequence obtained by sampling ECG signals, which will not be elaborated in the embodiments of the present application.

[0094] Step S202. Input the target ECG sequence into the ECG data processing model to perform ECG information mining to obtain heartbeat distribution information and event information of cardiac electrophysiological activities, wherein the heartbeat distribution information includes the heartbeat type in the target ECG sequence, and the event information includes the event coverage area and the event type; wherein the ECG data processing model is an end-to-end model that synchronously outputs the heartbeat type, the event coverage area and the event type.

[0095] The heartbeat in the embodiment of the present application refers to a complete process of cardiac contraction and relaxation, including ventricular contraction (pumping blood out of the heart) and relaxation (refilling the ventricle with blood). Under normal circumstances, the heartbeat is regular, and each heartbeat is accompanied by a heartbeat sound. The heartbeat frequency is usually referred to as the heart rate, i.e. the number of heartbeats per minute. The different behavioral manifestations of the heartbeat can be classified into several types. In an exemplary embodiment, the heartbeat type can be divided into 11 kinds, namely heartbeat O, normal heartbeat N, atrial premature beat S, ventricular premature beat V, bundle branch block B, ventricular preexcitation M, atrial escape beat Se, ventricular escape beat Ve, junctional escape beat Je, atrial premature beat Sa, and artifact X.

[0096] The event information of the cardiac electrophysiological activity in the embodiment of the present application points to the event generated by the cardiac electrophysiological activity. For the specific event type of the event generated by the cardiac electrophysiological activity, the embodiment of the present application is not limited. In an exemplary embodiment, the event can be divided into 24 kinds, namely sinus rhythm, sinus bradycardia, sinus tachycardia, single atrial premature beat, paired atrial premature beat, atrial premature beat bigeminy, atrial premature beat trigeminy, atrial tachycardia, single ventricular premature beat, paired ventricular premature beat, ventricular premature beat bigeminy, ventricular premature beat trigeminy, ventricular tachycardia, arrest, atrial fibrillation, atrial fibrillation with slow ventricular rate, atrial fibrillation with fast ventricular rate, ST segment elevation, ST segment depression, first degree atrioventricular block, second degree type I atrioventricular block, second degree type II atrioventricular block, third degree atrioventricular block, second degree type II sinoatrial block. The event coverage area refers to the area corresponding to the corresponding event in the target electrocardiogram sequence.

[0097] The ECG data processing method proposed in the embodiment of the present application is implemented through a trained end-to-end ECG data processing model, which synchronously outputs the heartbeat type, event coverage area and event type, so there is no need to combine other ECG signal analysis methods. For the input target ECG sequence, the heartbeat distribution information and event information of the cardiac electrophysiological activity in the target ECG sequence can be directly output end-to-end. Moreover, the ECG data processing model has a global vision, there is no limit on the length of the input target ECG sequence, and it does not need to be cut into ECG data segments of fixed length, and can be directly input into the ECG data processing model. The ECG data processing model directly mines the event information of the cardiac electrophysiological activity in an end-to-end mode, without relying on assumed information, greatly improving the accuracy of mining the event information of the cardiac electrophysiological activity.

[0098] In an exemplary embodiment, the method further includes: calibrating the heartbeat distribution information based on the event information to obtain calibrated heartbeat distribution information; or calibrating the event information based on the heartbeat distribution information to obtain calibrated event information. In other words, the event information and the heartbeat distribution information can each serve as a calibration information for the other to improve the accuracy of the other. This is because the event information and the heartbeat distribution information are simultaneously predicted by the end-to-end ECG data processing model and have homologous characteristics. Therefore, they can be further calibrated with each other to improve the accuracy of each other.

[0099] For example, in some cases, such as when there are events such as block, cardiac arrest, frequent atrial tachycardia, sinus tachycardia, and atrial fibrillation, it is difficult to distinguish heartbeats such as atrial premature beats S and atrial premature beats Sa. However, if the types and coverage areas of these events are used to assist in the heartbeat distinction, the difficulty of distinction can be significantly reduced, and the accuracy can be greatly improved. Of course, the embodiments of the present application do not limit the specific method of mutual calibration, and do not constitute an implementation obstacle.

[0100] Please refer to Figure 3 , which shows a flow chart of the ECG data processing model training method in an embodiment of the present application. The ECG data processing model is trained by the following method:

[0101] Step S301: Acquire a sample ECG sequence, and acquire heart beat reference information and event reference information corresponding to the sample ECG sequence.

[0102] For any sample ECG sequence, the sample ECG sequence can be divided into a plurality of sample ECG segments of equal length based on a first preset step length; each of the sample ECG segments can be divided into a plurality of sample ECG regions of equal length based on a second preset step length; wherein the heartbeat reference information includes a reference heartbeat type corresponding to each of the sample ECG regions; and the event reference information includes a reference event coverage area and a reference event type corresponding to each of the sample ECG segments. The embodiment of the present application does not limit the specific values ​​of the first preset step length and the second preset step length, but it is related to the structure of the ECG data processing model and cannot be set arbitrarily.

[0103] The reference heartbeat type indicates the specific heartbeat type of the corresponding sample ECG region. In an exemplary embodiment, the reference heartbeat type is represented by the following encoding:

[0104]

[0105]

[0106] The reference event coverage area indicates the continuous area of ​​the cardiac electrophysiological activity event occurring in the corresponding sample ECG segment, which is a kind of positioning information. The reference event type indicates the type of the cardiac electrophysiological activity event occurring in the corresponding sample ECG segment, which is a kind of type information.

[0107] The reference heartbeat type, reference event coverage area and reference event type are all based on the heartbeat types that have been marked in the sample ECG sequence, the heartbeat locations corresponding to the heartbeat types, the types and action areas of events of cardiac electrophysiological activities, and the reference information for each sample ECG segment and each sample ECG area. These reference information are used as true value data for model training.

[0108] Step S302: construct a preset model, which includes a global ECG feature extraction module, a local ECG feature extraction module, an ECG feature fusion module, an event location module, an event classification module, and a heart beat classification module.

[0109] The preset model is a model specially set in the embodiment of the present application for training the end-to-end ECG data processing model. Figure 4, which shows a schematic diagram of the architecture of the preset model in the embodiment of the present application. In the preset model, the global ECG feature extraction module 401 is connected to the local ECG feature extraction module 402, and the ECG feature fusion module 403 is used to splice the features output by the last network layer of the global ECG feature extraction module 401 with the features output by the last network layer of the local ECG feature extraction module 402, and perform feature fusion on the splicing result. The fusion results are transmitted to the event positioning module 404, the event classification module 405, and the heartbeat classification module 406 respectively, so as to perform event positioning, event classification and heartbeat classification respectively, so as to achieve the technical purpose of synchronously outputting the prediction results of event positioning (hereinafter referred to as the predicted event coverage area), the prediction results of event classification (hereinafter referred to as the predicted event type) and the prediction results of heartbeat classification (hereinafter referred to as the predicted heartbeat type).

[0110] Exemplarily, the global ECG feature extraction module 401 includes at least two convolutional layers for generating time scale variations; for example, the global ECG feature extraction module 401 includes 5 convolutional layers, and the step size of each convolutional layer is 2. Then, after passing through the global ECG feature extraction module 401, the data undergoes 5 dimensionality reductions, and the length is reduced by 32 times. In conjunction with this structure, the embodiment of the present application designs specific values ​​of the first preset step size and the second preset step size, as well as specific expressions of the heartbeat reference information and the event reference information, so as to achieve the purpose of realizing specific truth data construction and model training in conjunction with the preset model structure, and ultimately achieve end-to-end model training.

[0111] Exemplarily, the local ECG feature extraction module 402 includes at least two bidirectional long short-term memory network layers (BiLSTM). For example, the local ECG feature extraction module 402 may include three bidirectional long short-term memory network layers (BiLSTM layers), and the first Bi LSTM layer can process sequences and capture long-term dependencies in sequences, which is particularly important for ECG signal analysis. The second BiLSTM layer and the third Bi LSTM are used to enhance the model's ability to capture sequence features. Introducing several BiLSTM layers in the preset model can significantly improve the model's ability to integrate time series features. The Bi LSTM layer can more comprehensively capture the time series dependencies in the ECG signal by simultaneously analyzing past and future information, which is crucial for identifying and distinguishing different types of heartbeats and arrhythmias. This bidirectional information flow enables the model to provide richer feature representations and have a global perspective, thereby improving the accuracy of classification and reducing the false alarm rate. In addition, the bidirectional characteristics of the Bi LSTM layer help to improve the generalization ability of the model, so that it can maintain stable performance when facing new and unseen data. The BiLSTM layer can also improve the model’s robustness to noise, reducing the impact of noise through its memory units and locating the heartbeat more accurately.

[0112] Furthermore, in the ECG feature fusion layer, feature splicing can be performed, and the splicing result is subjected to feature fusion through a 1x1 convolution layer, which allows the model to recombine and fuse the channels of the features without changing the spatial dimension. This helps the model learn the correlation between different channels and enhances the expressiveness of the features. After feature fusion, the model synchronously passes the fusion results to the event location module 404, the event classification module 405, and the heartbeat classification module 406. This structural design has the following effects:

[0113] 1. Information sharing: By sharing the same high-dimensional features, the event location module 404, the event classification module 405, and the heartbeat classification module 406 can use the same information basis to perform their respective tasks. This helps to reduce repeated calculations and improve the efficiency of the model. The results output by the event location module 404, the event classification module 405, and the heartbeat classification module 406 are homologous, so they can support mutual verification.

[0114] 2. Feature complementarity: Event location module 404, event classification module 405, and heartbeat classification module 406 may extract different information from shared features. For example, event location module 404 may focus more on spatially related signals, while event classification module 405 and heartbeat classification module 406 may focus more on the statistical characteristics of the signals. This complementarity helps the model understand data from multiple perspectives and improves the comprehensiveness of the output.

[0115] 3. Improve learning ability: The event location module 404, the event classification module 405, and the heartbeat classification module 406 each learn from each other in learning, thereby achieving a partner learning effect and jointly improving learning ability. The model supports end-to-end training, which means that the model can optimize the parameters of all modules at the same time through the back propagation algorithm. This end-to-end learning helps to improve the overall performance of the model.

[0116] Step S303: Based on the global ECG feature extraction module, extract global ECG features from the sample ECG sequence to obtain global ECG features.

[0117] The embodiment of the present application can determine the first preset step size and the second preset step size based on the change in the time scale during the global ECG feature extraction process. Continuing with the previous example, the global ECG feature extraction module includes 5 convolutional layers, and the step size of each convolutional layer is 2. Then, after passing through the global ECG feature extraction module, the data undergoes 5 dimensionality reductions, and the length is reduced by 32 times.

[0118] In this case, in the sampled ECG sequence, each sampling point corresponds to an ECG signal, and the actual time represented by it is 2ms. Every 32,000 sampling points (i.e., the actual time is 64s) in the sample ECG sequence is used as the first preset step length to divide the sample ECG sequence to obtain a number of sample ECG segments, where the i-th sample ECG segment is recorded as Mi. If the last sample ECG segment is less than 32,000 sampling points, then zeros are added to the tail until 32,000 sampling points are reached, thereby obtaining multiple sample ECG segments of equal length. For each sample ECG segment Mi, there are 32,000 sampling points, and each sampling point represents an actual time of 2ms.

[0119] Since the 5-layer dimensionality reduction is reduced by 32 times in total, after the global ECG feature extraction, the actual time corresponding to each point is expanded from 2ms to 64ms. For Mi, the data points in CNN_Mi obtained after the global ECG feature extraction become 32000 / 32=1000, and each data point represents 64ms of actual time. Based on this, each sample ECG segment Mi of 32,000 sampling points can be further divided into 1000 (second preset step size) sample ECG regions Bi, each sample ECG region Bi contains 32 points. After the global ECG feature extraction, each sample ECG region Bi has only one point, which is equivalent to the original 32 sampling points.

[0120] Step S304. Based on the local ECG feature extraction module, perform local ECG feature extraction on the global ECG feature to obtain local ECG features; input the global ECG features and the local ECG features into the ECG feature fusion module to perform ECG feature fusion to obtain fused ECG features; input the fused ECG features into the heart beat classification module, the event location module and the event classification module respectively to perform heart beat classification, event location and event classification respectively to obtain sample heart beat distribution information and sample event information of cardiac electrophysiological activities.

[0121] Specifically, the sample heartbeat distribution information includes the predicted heartbeat type corresponding to each sample ECG region, and the sample event information includes the predicted event coverage area and predicted event type corresponding to each sample ECG segment.

[0122] Step S305: Determine a first loss based on the difference between the sample heartbeat distribution information and the heartbeat reference information.

[0123] Please refer to Figure 5 , which shows a schematic flow chart of a method for determining a first loss in an embodiment of the present application. Specifically, determining the first loss based on the difference between the sample heartbeat distribution information and the heartbeat reference information includes:

[0124] Step S501. For any of the sample ECG regions, based on the difference between a reference heartbeat type corresponding to the sample ECG region and a corresponding predicted heartbeat type, determine a corresponding heartbeat type prediction loss.

[0125] The embodiment of the present application does not limit the quantification method of the difference between the reference heartbeat type corresponding to the sample ECG region and the corresponding predicted heartbeat type. For example, cross entropy loss may be used.

[0126] Step S502: Determine the first loss based on the predicted loss of the heartbeat types corresponding to the sample electrocardiogram regions.

[0127] Exemplarily, the sum of the heartbeat type prediction losses corresponding to each of the sample electrocardiogram regions may be determined as the first loss.

[0128] Step S306: Determine a second loss based on the difference between the sample event information and the event reference information.

[0129] Please refer to Figure 6 , which shows a schematic flow chart of a method for determining a second loss in an embodiment of the present application. Specifically, determining the second loss based on the difference between the sample event information and the event reference information includes:

[0130] Step S601. For any of the sample ECG segments, based on the difference between the reference event coverage area corresponding to the sample ECG segment and the corresponding predicted event coverage area, determine the corresponding event positioning loss; and based on the difference between the reference event type corresponding to the sample ECG segment and the corresponding predicted event type, determine the corresponding event type prediction loss.

[0131] The embodiment of the present application does not limit the quantification method of the difference between the reference event coverage area corresponding to the sample ECG segment and the corresponding predicted event coverage area, for example, the mean absolute error loss can be used. Of course, the quantification method of the difference between the reference event type corresponding to the sample ECG segment and the corresponding predicted event type is also not limited, for example, the cross entropy loss can be used.

[0132] Step S602: Determine the second loss based on the event localization loss corresponding to each of the sample ECG segments and the event type prediction loss corresponding to each of the sample ECG segments.

[0133] Exemplarily, the sum of the event localization losses corresponding to each of the sample ECG segments and the event type prediction losses corresponding to each of the sample ECG segments can be determined as the second loss.

[0134] Step S307: Based on the first loss and the second loss, adjust the parameters of each module in the preset model to obtain the ECG data processing model.

[0135] Specifically, the weighted sum of the first loss and the second loss can be used as the total loss of the model. Of course, the embodiment of the present application does not limit the weight value, which can be set according to actual conditions. Based on the total loss, the parameters of each module in the preset model are adjusted, and after the parameter adjustment stops, the ECG data processing model is obtained.

[0136] The embodiment of the present application does not limit the parameter adjustment method. For example, the gradient descent method can be used. When the number of parameter adjustments reaches the preset number threshold, or when the total loss is less than the preset loss threshold, the parameter adjustment can be stopped to obtain the ECG data processing model. The embodiment of the present application does not set the preset number threshold and the preset loss threshold, which can be freely selected according to the actual situation and does not constitute an implementation obstacle.

[0137] The following is an embodiment of the device of the present application, which can be used to execute the embodiment of the method of the present application. For details not disclosed in the embodiment of the device of the present application, please refer to the embodiment of the method of the present application.

[0138] Please refer to Figure 7, which shows a block diagram of an ECG data processing device of an exemplary embodiment. The device has the function of implementing the above ECG data processing method, and the above function can be implemented by hardware, or by hardware executing corresponding software. The device can be a computer device, or can be set in a computer device. The device may include:

[0139] A target ECG sequence acquisition module 701 is used to acquire a target ECG sequence;

[0140] The target ECG sequence processing module 702 is used to input the target ECG sequence into the ECG data processing model to perform ECG information mining to obtain heartbeat distribution information and event information of cardiac electrophysiological activities, wherein the heartbeat distribution information includes the heartbeat type in the target ECG sequence, and the event information includes the event coverage area and event type;

[0141] The electrocardiogram data processing model is an end-to-end model that synchronously outputs the heartbeat type, the event coverage area, and the event type.

[0142] In an exemplary embodiment, the target ECG sequence processing module 702 is configured to perform the following operations:

[0143] Based on the event information, calibrating the heartbeat distribution information to obtain calibrated heartbeat distribution information;

[0144] or,

[0145] Based on the heartbeat distribution information, the event information is calibrated to obtain calibrated event information.

[0146] In an exemplary embodiment, the target ECG sequence processing module 702 is configured to perform the following operations:

[0147] Acquire a sample ECG sequence, and acquire heart beat reference information and event reference information corresponding to the sample ECG sequence;

[0148] Constructing a preset model, wherein the preset model includes a global ECG feature extraction module, a local ECG feature extraction module, an ECG feature fusion module, an event location module, an event classification module, and a heart beat classification module;

[0149] Based on the global ECG feature extraction module, extract global ECG features from the sample ECG sequence to obtain global ECG features;

[0150] Based on the local ECG feature extraction module, extracting local ECG features from the global ECG features to obtain local ECG features;

[0151] Inputting the global ECG feature and the local ECG feature into the ECG feature fusion module to perform ECG feature fusion to obtain a fused ECG feature;

[0152] Inputting the fused ECG features into the heartbeat classification module, the event location module and the event classification module respectively, performing heartbeat classification, event location and event classification respectively, and obtaining sample heartbeat distribution information and sample event information of cardiac electrophysiological activities;

[0153] determining a first loss based on a difference between the sample heartbeat distribution information and the heartbeat reference information;

[0154] determining a second loss based on a difference between the sample event information and the event reference information;

[0155] Based on the first loss and the second loss, the parameters of each module in the preset model are adjusted to obtain the electrocardiogram data processing model.

[0156] In an exemplary embodiment, the target ECG sequence processing module 702 is configured to perform the following operations:

[0157] Dividing the sample ECG sequence into a plurality of sample ECG segments of equal length based on a first preset step length;

[0158] Dividing each of the sample ECG segments into a plurality of sample ECG regions of equal length based on a second preset step length;

[0159] The heartbeat reference information includes a reference heartbeat type corresponding to each of the sample ECG regions; and the event reference information includes a reference event coverage area and a reference event type corresponding to each of the sample ECG segments.

[0160] In an exemplary embodiment, the target ECG sequence processing module 702 is configured to perform the following operations:

[0161] The first preset step size and the second preset step size are determined based on the change in the time scale during the global electrocardiogram feature extraction process.

[0162] In an exemplary embodiment, the sample heartbeat distribution information includes the predicted heartbeat type corresponding to each sample ECG region. The target ECG sequence processing module 702 is used to perform the following operations:

[0163] For any of the sample ECG regions, determining a corresponding heartbeat type prediction loss based on a difference between a reference heartbeat type corresponding to the sample ECG region and a corresponding predicted heartbeat type;

[0164] The first loss is determined based on the predicted loss of the heartbeat types corresponding to the sample electrocardiogram regions.

[0165] In an exemplary embodiment, the sample event information includes the predicted event coverage area and predicted event type corresponding to each sample ECG segment, and the target ECG sequence processing module 702 is used to perform the following operations:

[0166] For any of the sample ECG segments, based on the difference between the reference event coverage area corresponding to the sample ECG segment and the corresponding predicted event coverage area, determine the corresponding event positioning loss; and based on the difference between the reference event type corresponding to the sample ECG segment and the corresponding predicted event type, determine the corresponding event type prediction loss;

[0167] The second loss is determined based on the event localization losses respectively corresponding to the sample ECG segments and the event type prediction losses respectively corresponding to the sample ECG segments.

[0168] In an exemplary embodiment, the global ECG feature extraction module includes at least two convolutional layers for generating a time-scale variation;

[0169] The local ECG feature extraction module includes at least two bidirectional long short-term memory network layers;

[0170] The ECG feature fusion module is used to splice the features output by the last network layer of the global ECG feature extraction module with the features output by the last network layer of the local ECG feature extraction module, and perform feature fusion on the splicing result.

[0171] It should be noted that the device provided in the above embodiment, when implementing its functions, is only illustrated by the division of the above functional modules. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the device and method embodiments provided in the above embodiment belong to the same concept, and the specific implementation process is detailed in the method embodiment, which will not be repeated here.

[0172] Please refer to Figure 8 , which shows a block diagram of a computer device according to an exemplary embodiment, for executing the above-mentioned ECG data processing method. Specifically:

[0173] The computer device 800 includes a central processing unit (CPU) 801, a system memory 804 including a random access memory (RAM) 802 and a read-only memory (ROM) 803, and a system bus 805 connecting the system memory 804 and the central processing unit 801. The computer device 800 also includes a basic input / output system (I / O (Input / Output) system) 806 that helps to transmit information between various components in the computer, and a large-capacity storage device 807 for storing an operating system 813, application programs 814 and other program modules 815.

[0174] The basic input / output system 806 includes a display 808 for displaying information and an input device 809 such as a mouse and a keyboard for user inputting information. The display 808 and the input device 809 are connected to the central processing unit 801 through an input / output controller 810 connected to the system bus 805. The basic input / output system 806 may also include an input / output controller 810 for receiving and processing inputs from a plurality of other devices such as a keyboard, a mouse, or an electronic stylus. Similarly, the input / output controller 810 also provides output to a display screen, a printer, or other types of output devices.

[0175] The mass storage device 807 is connected to the central processing unit 801 through a mass storage controller (not shown) connected to the system bus 805. The mass storage device 807 and its associated computer readable medium provide non-volatile storage for the computer device 800. That is, the mass storage device 807 may include a computer readable medium (not shown) such as a hard disk or a CD-ROM (Compact Disk Read-Only Memory) drive.

[0176] Without loss of generality, computer readable media may include computer storage media and communication media. Computer storage media include volatile and non-volatile, removable and non-removable media implemented by any method or technology for storing information such as computer readable instructions, data structures, program modules or other data. Computer storage media include RAM, ROM, EPROM (Erasable Programmable Read On ly Memory), EEPROM (Electrically Erasable Programmable Read On ly Memory), flash memory or other solid-state storage technologies, CD-ROM, DVD (Digital Video Disc) or other optical storage, cassettes, magnetic tapes, disk storage or other magnetic storage devices. Of course, those skilled in the art will appreciate that computer storage media are not limited to the above. The above-mentioned system memory 804 and mass storage device 807 can be collectively referred to as memory.

[0177] According to various embodiments of the present application, the computer device 800 can also be connected to a remote computer on the network through a network such as the Internet. That is, the computer device 800 can be connected to the network 812 through the network interface unit 811 connected to the system bus 805, or the network interface unit 811 can be used to connect to other types of networks or remote computer systems (not shown).

[0178] The memory also includes a computer program, which is stored in the memory and configured to be executed by one or more processors to implement the ECG data processing method.

[0179] In an exemplary embodiment, a computer-readable storage medium is also provided, in which at least one instruction, at least one program, a code set or an instruction set is stored. When the at least one instruction, the at least one program, the code set or the instruction set is executed by a processor, the above-mentioned electrocardiogram data processing method is implemented.

[0180] Specifically, the electrocardiogram data processing method includes:

[0181] Acquire target ECG sequence;

[0182] Inputting the target ECG sequence into an ECG data processing model to perform ECG information mining to obtain heartbeat distribution information and event information of cardiac electrophysiological activities, wherein the heartbeat distribution information includes the heartbeat type in the target ECG sequence, and the event information includes the event coverage area and the event type;

[0183] The electrocardiogram data processing model is an end-to-end model that synchronously outputs the heartbeat type, the event coverage area, and the event type.

[0184] In an exemplary embodiment, the method further comprises:

[0185] Based on the event information, calibrating the heartbeat distribution information to obtain calibrated heartbeat distribution information;

[0186] or,

[0187] Based on the heartbeat distribution information, the event information is calibrated to obtain calibrated event information.

[0188] In an exemplary embodiment, the ECG data processing model is trained by the following method:

[0189] Acquire a sample ECG sequence, and acquire heart beat reference information and event reference information corresponding to the sample ECG sequence;

[0190] Constructing a preset model, wherein the preset model includes a global ECG feature extraction module, a local ECG feature extraction module, an ECG feature fusion module, an event location module, an event classification module, and a heart beat classification module;

[0191] Based on the global ECG feature extraction module, extract global ECG features from the sample ECG sequence to obtain global ECG features;

[0192] Based on the local ECG feature extraction module, extracting local ECG features from the global ECG features to obtain local ECG features;

[0193] Inputting the global ECG feature and the local ECG feature into the ECG feature fusion module to perform ECG feature fusion to obtain a fused ECG feature;

[0194] Inputting the fused ECG features into the heartbeat classification module, the event location module and the event classification module respectively, performing heartbeat classification, event location and event classification respectively, and obtaining sample heartbeat distribution information and sample event information of cardiac electrophysiological activities;

[0195] determining a first loss based on a difference between the sample heartbeat distribution information and the heartbeat reference information;

[0196] determining a second loss based on a difference between the sample event information and the event reference information;

[0197] Based on the first loss and the second loss, the parameters of each module in the preset model are adjusted to obtain the electrocardiogram data processing model.

[0198] In an exemplary embodiment, the method further comprises:

[0199] Dividing the sample ECG sequence into a plurality of sample ECG segments of equal length based on a first preset step length;

[0200] Dividing each of the sample ECG segments into a plurality of sample ECG regions of equal length based on a second preset step length;

[0201] The heartbeat reference information includes a reference heartbeat type corresponding to each of the sample ECG regions; and the event reference information includes a reference event coverage area and a reference event type corresponding to each of the sample ECG segments.

[0202] In an exemplary embodiment, the method further comprises:

[0203] The first preset step size and the second preset step size are determined based on the change in the time scale during the global electrocardiogram feature extraction process.

[0204] In an exemplary embodiment, the sample heartbeat distribution information includes a predicted heartbeat type corresponding to each of the sample electrocardiogram regions, and determining the first loss based on a difference between the sample heartbeat distribution information and the heartbeat reference information includes:

[0205] For any of the sample ECG regions, determining a corresponding heartbeat type prediction loss based on a difference between a reference heartbeat type corresponding to the sample ECG region and a corresponding predicted heartbeat type;

[0206] The first loss is determined based on the predicted loss of the heartbeat types corresponding to the sample electrocardiogram regions.

[0207] In an exemplary embodiment, the sample event information includes a predicted event coverage area and a predicted event type corresponding to each of the sample ECG segments, and determining the second loss based on the difference between the sample event information and the event reference information includes:

[0208] For any of the sample ECG segments, based on the difference between the reference event coverage area corresponding to the sample ECG segment and the corresponding predicted event coverage area, determine the corresponding event positioning loss; and based on the difference between the reference event type corresponding to the sample ECG segment and the corresponding predicted event type, determine the corresponding event type prediction loss;

[0209] The second loss is determined based on the event localization losses respectively corresponding to the sample ECG segments and the event type prediction losses respectively corresponding to the sample ECG segments.

[0210] In an exemplary embodiment, the global ECG feature extraction module includes at least two convolutional layers for generating a time-scale variation;

[0211] The local ECG feature extraction module includes at least two bidirectional long short-term memory network layers;

[0212] The ECG feature fusion module is used to splice the features output by the last network layer of the global ECG feature extraction module with the features output by the last network layer of the local ECG feature extraction module, and perform feature fusion on the splicing result.

[0213] Optionally, the computer readable storage medium may include: ROM (Read On ly Memory), RAM (Random Access Memory), SSD (Solid State Drives) or an optical disk, etc. Among them, the random access memory may include ReRAM (Resistance Random Access Memory) and DRAM (Dynamic Random Access Memory).

[0214] In an exemplary embodiment, a computer program product or a computer program is also provided, the computer program product or the computer program comprising computer instructions, the computer instructions being stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the above-mentioned electrocardiogram data processing method.

[0215] It should be understood that the "multiple" mentioned in this article refers to two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the objects associated before and after are in an "or" relationship. In addition, the step numbers described in this article only illustrate a possible execution sequence between the steps. In some other embodiments, the above steps may not be executed in the order of the numbers, such as two steps with different numbers are executed at the same time, or two steps with different numbers are executed in the opposite order to the diagram. The embodiments of the present application are not limited to this.

[0216] In addition, in the specific implementation of this application, related data such as user information is involved. When the above embodiments of this application are applied to specific products or technologies, user permission or consent is required, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards of relevant countries and regions.

[0217] The above are merely exemplary embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method for processing electrocardiogram data, characterized in that: The method comprises: Acquire target ECG sequence; Inputting the target ECG sequence into an ECG data processing model to perform ECG information mining to obtain heartbeat distribution information and event information of cardiac electrophysiological activities, wherein the heartbeat distribution information includes the heartbeat type in the target ECG sequence, and the event information includes event coverage area and event type; The electrocardiogram data processing model is an end-to-end model that synchronously outputs the heartbeat type, the event coverage area, and the event type.

2. The method according to claim 1, characterized in that The method further comprises: Based on the event information, calibrating the heartbeat distribution information to obtain calibrated heartbeat distribution information; or, Based on the heartbeat distribution information, the event information is calibrated to obtain calibrated event information.

3. The method according to claim 1 or 2, characterized in that: The ECG data processing model is trained by the following method: Acquire a sample ECG sequence, and acquire heart beat reference information and event reference information corresponding to the sample ECG sequence; Constructing a preset model, wherein the preset model includes a global ECG feature extraction module, a local ECG feature extraction module, an ECG feature fusion module, an event positioning module, an event classification module, and a heart beat classification module; Based on the global ECG feature extraction module, extract global ECG features from the sample ECG sequence to obtain global ECG features; Based on the local ECG feature extraction module, extracting local ECG features from the global ECG features to obtain local ECG features; Inputting the global ECG feature and the local ECG feature into the ECG feature fusion module to perform ECG feature fusion to obtain a fused ECG feature; Inputting the fused ECG features into the heartbeat classification module, the event location module and the event classification module respectively, performing heartbeat classification, event location and event classification respectively, and obtaining sample heartbeat distribution information and sample event information of cardiac electrophysiological activities; determining a first loss based on a difference between the sample heartbeat distribution information and the heartbeat reference information; determining a second loss based on a difference between the sample event information and the event reference information; Based on the first loss and the second loss, the parameters of each module in the preset model are adjusted to obtain the electrocardiogram data processing model.

4. The method according to claim 3, characterized in that The method further comprises: Dividing the sample ECG sequence into a plurality of sample ECG segments of equal length based on a first preset step length; Dividing each of the sample ECG segments into a plurality of sample ECG regions of equal length based on a second preset step length; The heartbeat reference information includes a reference heartbeat type corresponding to each of the sample ECG regions; and the event reference information includes a reference event coverage area and a reference event type corresponding to each of the sample ECG segments.

5. The method according to claim 4, characterized in that The method further comprises: The first preset step size and the second preset step size are determined based on the change in the time scale during the global electrocardiogram feature extraction process.

6. The method according to claim 5, characterized in that The sample heartbeat distribution information includes a predicted heartbeat type corresponding to each sample electrocardiogram region, and determining the first loss based on a difference between the sample heartbeat distribution information and the heartbeat reference information includes: For any of the sample ECG regions, determining a corresponding heartbeat type prediction loss based on a difference between a reference heartbeat type corresponding to the sample ECG region and a corresponding predicted heartbeat type; The first loss is determined based on the predicted loss of the heartbeat types corresponding to the sample electrocardiogram regions.

7. The method according to claim 6, characterized in that The sample event information includes a predicted event coverage area and a predicted event type corresponding to each sample ECG segment, and determining the second loss based on a difference between the sample event information and the event reference information includes: For any of the sample ECG segments, based on the difference between the reference event coverage area corresponding to the sample ECG segment and the corresponding predicted event coverage area, determine the corresponding event positioning loss; and based on the difference between the reference event type corresponding to the sample ECG segment and the corresponding predicted event type, determine the corresponding event type prediction loss; The second loss is determined based on the event localization losses respectively corresponding to the sample ECG segments and the event type prediction losses respectively corresponding to the sample ECG segments.

8. The method according to claim 7, characterized in that The global ECG feature extraction module includes at least two convolutional layers for generating a time scale variation; The local ECG feature extraction module includes at least two bidirectional long short-term memory network layers; The ECG feature fusion module is used to splice the features output by the last network layer of the global ECG feature extraction module with the features output by the last network layer of the local ECG feature extraction module, and perform feature fusion on the splicing result.

9. An electrocardiogram data processing device, characterized in that: The device comprises: A target ECG sequence acquisition module is used to acquire a target ECG sequence; A target ECG sequence processing module is used to input the target ECG sequence into an ECG data processing model to perform ECG information mining to obtain heartbeat distribution information and event information of cardiac electrophysiological activities, wherein the heartbeat distribution information includes the heartbeat type in the target ECG sequence, and the event information includes event coverage area and event type; The electrocardiogram data processing model is an end-to-end model that synchronously outputs the heartbeat type, the event coverage area, and the event type.

10. A computer-readable storage medium, characterized in that: The storage medium stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by the processor to implement the ECG data processing method as described in any one of claims 1 to 8.

Images