Electrocardiogram multi-task analysis method, device and equipment and storage medium

Abstract

This application relates to the field of electrocardiogram (ECG) signal processing technology, and in particular to a method, apparatus, device, and storage medium for ECG multi-task analysis. The method includes: acquiring an ECG signal; inputting the ECG signal into a feature extraction network for feature extraction to obtain ECG signal features; inputting the ECG signal features into deep learning networks corresponding to heartbeat localization, reference point detection, and heartbeat classification for calculation to obtain scores and uncertainties corresponding to heartbeat localization, reference point detection, and heartbeat classification; and calculating the heartbeat localization result, reference point detection result, and heartbeat classification result based on the heartbeat localization score, reference point detection score, heartbeat classification score, heartbeat localization uncertainty, reference point detection uncertainty, and heartbeat classification uncertainty. The method of this application can solve the technical problem of existing ECG multi-task analysis methods struggling to balance computational load and accuracy when deployed on platforms with limited computing resources.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electrocardiosignal processing, and particularly relates to an electrocardio multi-task analysis method, device, equipment and storage medium. BACKGROUND

[0002] As an important cardiovascular state examination means, electrocardiogram (ECG) is used more and more on wearable devices (such as smart watches). Compared with professional medical devices such as electrocardiograph, monitor, Holter, etc., the computing resources of wearable devices are quite limited due to power consumption, size, cost and complex use scenarios, and the collected electrocardiosignal is often disturbed by noise. Heartbeat positioning, fiducial point detection and heartbeat classification are three important tasks of electrocardio analysis, and how to accurately, efficiently and reliably complete these tasks on wearable devices is a challenge.

[0003] Traditional electrocardio analysis methods based on rules or machine learning have small calculation amount, but their accuracy is relatively low compared with deep learning methods due to the limitation of model expression ability. Deep learning methods have high accuracy, but most of them use multiple single-task networks, that is, multiple corresponding independent feature extraction networks are used to analyze electrocardio multi-task, which has the disadvantage of too large calculation amount. Therefore, some methods use a single single-task network, that is, an electrocardio multi-task analysis is performed through a feature extraction network, sacrificing some accuracy to save a large amount of computing resources. However, due to the uneven quality of the input ECG signal, it cannot be used for all analysis tasks, for example, the signal quality of a certain segment of ECG is poor, which can only be used for heartbeat positioning, and cannot be used for fiducial point detection and heartbeat classification. If the ECG signal with poor quality is forced to be used for analysis, the analysis result will be unreliable.

[0004] Therefore, the existing electrocardio multi-task analysis method is difficult to balance the calculation amount and accuracy when deployed in a platform with limited computing resources. SUMMARY

[0005] The main purpose of the present application is to provide an electrocardio multi-task analysis method, device, equipment and storage medium, which aims to solve the technical problem that the existing electrocardio multi-task analysis method is difficult to balance the calculation amount and accuracy when deployed in a platform with limited computing resources.

[0006] In order to achieve the above-mentioned purpose of the application, the present application provides an electrocardio multi-task analysis method, which comprises:

[0007] obtaining an electrocardiosignal;

[0008] inputting the electrocardiosignal into a feature extraction network for feature extraction to obtain an electrocardiosignal feature;

[0009] The ECG signal features are input into the deep learning networks corresponding to heartbeat localization, reference point detection, and heartbeat classification, respectively, to calculate the heartbeat localization score, reference point detection score, heartbeat classification score, heartbeat localization uncertainty, reference point detection uncertainty, and heartbeat classification uncertainty.

[0010] The heartbeat positioning score, the reference point detection score, the heartbeat classification score, the heartbeat positioning uncertainty, the reference point detection uncertainty, and the heartbeat classification uncertainty are calculated to obtain the heartbeat positioning result, the reference point detection result, and the heartbeat classification result.

[0011] Further, the step of inputting the ECG signal features into the deep learning networks corresponding to heartbeat localization, reference point detection, and heartbeat classification respectively for calculation to obtain the heartbeat localization score, reference point detection score, heartbeat classification score, heartbeat localization uncertainty, reference point detection uncertainty, and heartbeat classification uncertainty includes:

[0012] The ECG signal features are input into the heartbeat localization network for calculation to obtain the heartbeat localization score and heartbeat localization uncertainty.

[0013] The ECG signal features are input into the reference point detection network for calculation to obtain the reference point detection score and the reference point detection uncertainty.

[0014] The ECG signal features are input into a heartbeat classification network for calculation to obtain the heartbeat classification score and heartbeat classification uncertainty.

[0015] Further, the step of calculating the heartbeat positioning result, the reference point detection result, and the heartbeat classification result based on the heartbeat positioning score, the reference point detection score, the heartbeat classification score, the heartbeat positioning uncertainty, the reference point detection uncertainty, and the heartbeat classification uncertainty includes:

[0016] The heartbeat positioning result is obtained by calculating based on the heartbeat positioning score and the heartbeat positioning uncertainty;

[0017] The reference point detection result is obtained by calculating the reference point detection score, the heartbeat positioning uncertainty, and the reference point detection uncertainty.

[0018] The heartbeat classification result is obtained by calculating the heartbeat classification score, the heartbeat positioning uncertainty, the reference point detection uncertainty, and the heartbeat classification uncertainty.

[0019] Further, the step of calculating the heartbeat positioning result based on the heartbeat positioning score and the heartbeat positioning uncertainty includes:

[0020] The heartbeat localization result is calculated using the following formula:

[0021]

[0022] Where, p loc For the heart rate localization results, s loc For heartbeat positioning score, u loc Let represent the uncertainty in heartbeat positioning, and let softmax be the normalized exponential function.

[0023] Further, the step of calculating the reference point detection result based on the reference point detection score, the heartbeat positioning uncertainty, and the reference point detection uncertainty includes:

[0024] The benchmark detection results are calculated using the following formula:

[0025]

[0026] Where, p del As the benchmark detection result, s del The benchmark detection score, u loc For the uncertainty of heartbeat positioning, u del The benchmark is used to determine the uncertainty, and the softmax function is a normalized exponential function.

[0027] Further, the step of calculating the heartbeat classification result based on the heartbeat classification score, the heartbeat positioning uncertainty, the reference point detection uncertainty, and the heartbeat classification uncertainty includes:

[0028] The heart rate classification result is calculated using the following formula:

[0029]

[0030] Where, p cls For the heart-shot classification results, s cls Classify the heartbeats into scores, u cls For the classification uncertainty of heartbeats, u loc For the uncertainty of heartbeat positioning, u del The benchmark is used to determine the uncertainty, and the softmax function is a normalized exponential function.

[0031] This application also provides an electrocardiogram multitasking analysis device, the device comprising:

[0032] The acquisition module is used to acquire electrocardiogram (ECG) signals;

[0033] The feature extraction module is used to input the electrocardiogram (ECG) signal into the feature extraction network for feature extraction to obtain ECG signal features;

[0034] The multi-task network computing module is used to input the ECG signal features into the deep learning networks corresponding to heartbeat localization, reference point detection, and heartbeat classification respectively for calculation, and obtain heartbeat localization score, reference point detection score, heartbeat classification score, heartbeat localization uncertainty, reference point detection uncertainty, and heartbeat classification uncertainty;

[0035] The electrocardiogram (ECG) analysis module is used to calculate, based on the heartbeat localization score, the reference point detection score, the heartbeat classification score, the heartbeat localization uncertainty, the reference point detection uncertainty, and the heartbeat classification uncertainty, to obtain the heartbeat localization result, the reference point detection result, and the heartbeat classification result.

[0036] Furthermore, the ECG multitasking analysis device includes:

[0037] The multi-task network calculation submodule is used to input the ECG signal features into the heartbeat localization network for calculation to obtain the heartbeat localization score and heartbeat localization uncertainty; input the ECG signal features into the reference point detection network for calculation to obtain the reference point detection score and reference point detection uncertainty; and input the ECG signal features into the heartbeat classification network for calculation to obtain the heartbeat classification score and heartbeat classification uncertainty.

[0038] This application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of any of the methods described above.

[0039] This application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the methods described above.

[0040] This application provides a multi-task ECG analysis method. It extracts ECG signal features using a feature extraction network, and then inputs these features into deep learning networks corresponding to multiple ECG analysis tasks to obtain scores and uncertainties for each task. This method can improve analysis accuracy while reducing computational load. Based on the different noise sensitivities of each task, the uncertainty of each task is adjusted during the calculation of the analysis results. This ensures that tasks with high noise sensitivity are less affected by noise samples, while tasks with low noise sensitivity avoid losing too many samples, further improving the accuracy of ECG analysis. Attached Figure Description

[0041] Figure 1 This is a flowchart illustrating an embodiment of the ECG multitasking analysis method of this application;

[0042] Figure 2This is a schematic block diagram of the structure of an electrocardiogram multitasking analysis device according to an embodiment of this application;

[0043] Figure 3 This is a schematic block diagram of the structure of a computer device according to an embodiment of this application.

[0044] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0045] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0046] Those skilled in the art will understand that, unless specifically stated otherwise, the singular forms “a,” “an,” “the,” and “the” used herein may also include the plural forms. It should be further understood that the term “comprising” as used in this specification means the presence of features, integers, steps, operations, elements, modules, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, modules, components, and / or groups thereof. It should be understood that when we say an element is “connected” or “coupled” to another element, it can be directly connected or coupled to the other element, or there may be intermediate elements. Furthermore, “connected” or “coupled” as used herein can include wireless connections or wireless coupling. The term “and / or” as used herein includes all or any modules and all combinations of one or more associated listed items.

[0047] It will be understood by those skilled in the art that, unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. It should also be understood that terms such as those defined in general dictionaries should be understood to have the same meaning as in the context of the prior art, and should not be interpreted in an idealized or overly formal sense unless specifically defined as herein.

[0048] Reference Figure 1 This invention provides a method for multitasking electrocardiogram analysis, including steps S1-S4, specifically:

[0049] S1. Obtain electrocardiogram (ECG) signals;

[0050] S2. Input the electrocardiogram (ECG) signal into a feature extraction network to extract features and obtain ECG signal features;

[0051] The ECG signal's amplitude, peak value, waveform index, and power spectral density are extracted using a feature extraction network. These extracted ECG features are then input into the deep learning networks corresponding to heartbeat localization, reference point detection, and heartbeat classification for computation. By inputting the same ECG signal features into multiple single-task networks, the number of steps involved in feature extraction within each network is reduced, thus decreasing computational load. Multiple single-task networks can independently and in parallel calculate the scores and uncertainties for their respective ECG tasks, improving the efficiency of multi-task ECG analysis.

[0052] S3. Input the ECG signal features into the deep learning networks corresponding to heartbeat localization, reference point detection, and heartbeat classification respectively for calculation to obtain heartbeat localization score, reference point detection score, heartbeat classification score, heartbeat localization uncertainty, reference point detection uncertainty, and heartbeat classification uncertainty;

[0053] S4. Calculate the heartbeat positioning score, the reference point detection score, the heartbeat classification score, the heartbeat positioning uncertainty, the reference point detection uncertainty, and the heartbeat classification uncertainty to obtain the heartbeat positioning result, the reference point detection result, and the heartbeat classification result.

[0054] ECG signal features are input into deep learning networks corresponding to heartbeat localization, reference point detection, and heartbeat classification, respectively, yielding scores and uncertainties for each task. By setting up multiple single-task networks to handle the corresponding ECG analysis tasks independently and in parallel, the accuracy and efficiency of ECG analysis can be improved. Based on the noise sensitivity of each task (heartbeat localization, reference point detection, and heartbeat classification), the uncertainties for each task are adjusted during the analysis calculation, making the output uncertainty more consistent with prior knowledge, thus improving the accuracy of each task's analysis and further enhancing the overall accuracy of ECG analysis.

[0055] As described above, features of the electrocardiogram (ECG) signal are extracted using a feature extraction network. These features are then input into deep learning networks corresponding to multiple ECG analysis tasks, yielding scores and uncertainties for each task. This approach reduces computational load while improving analysis accuracy. Furthermore, based on the different noise sensitivities of each task, the uncertainty of each task is adjusted during the calculation of analysis results. This ensures that tasks with high noise sensitivity are less affected by noisy samples, while tasks with low noise sensitivity avoid losing too many samples, further improving the accuracy of ECG analysis.

[0056] In one embodiment, the steps of inputting the ECG signal features into the deep learning networks corresponding to heartbeat localization, reference point detection, and heartbeat classification for calculation to obtain heartbeat localization score, reference point detection score, heartbeat classification score, heartbeat localization uncertainty, reference point detection uncertainty, and heartbeat classification uncertainty include:

[0057] S301. Input the ECG signal features into the heartbeat localization network for calculation to obtain the heartbeat localization score and heartbeat localization uncertainty.

[0058] S302. Input the ECG signal features into the reference point detection network for calculation to obtain the reference point detection score and the reference point detection uncertainty;

[0059] S303. Input the ECG signal features into the heartbeat classification network for calculation to obtain the heartbeat classification score and heartbeat classification uncertainty.

[0060] When setting the output of the heartbeat localization network, channel 0 can be set to represent non-heartbeats, and channel 1 to represent heartbeats. When setting the output of the benchmark detection network, channel 0 can be set to represent non-characteristic waves, channel 1 to represent P waves, channel 2 to represent QRS waves, and channel 3 to represent T waves. When setting the output of the heartbeat classification network, channel 0 can be set to represent non-heartbeats, channel 1 to represent sinus beats, channel 2 to represent supraventricular beats, channel 3 to represent ventricular beats, channel 4 to represent fused beats, and channel 5 to represent other types of beats, such as pacemaker beats. The network structures of the heartbeat localization network, benchmark detection network, and heartbeat classification network can be modified according to actual needs. Uncertainty evaluation modules for the corresponding tasks have been added to the heartbeat localization network, benchmark detection network, and heartbeat classification network, and the output uncertainty is accumulated based on the task's noise sensitivity. When applied to the training of deep learning networks for multi-task tasks, it can reduce the interference of noisy samples on tasks with high signal quality requirements, while allowing tasks with low signal quality requirements to make full use of the training data. In actual ECG multi-task analysis, deep learning networks for multi-task tasks can output the uncertainty of each task to evaluate the reliability of the analysis results.

[0061] In one embodiment, the steps of calculating the heartbeat positioning result, the reference point detection result, and the heartbeat classification result based on the heartbeat positioning score, the reference point detection score, the heartbeat classification score, the heartbeat positioning uncertainty, the reference point detection uncertainty, and the heartbeat classification uncertainty include:

[0062] S401. Calculate the heartbeat positioning result based on the heartbeat positioning score and the heartbeat positioning uncertainty;

[0063] S402. Calculate the reference point detection result based on the reference point detection score, the heartbeat positioning uncertainty, and the reference point detection uncertainty;

[0064] S403. Calculate the heartbeat classification result based on the heartbeat classification score, the heartbeat positioning uncertainty, the reference point detection uncertainty, and the heartbeat classification uncertainty.

[0065] Since the three tasks—heartbeat localization, reference point detection, and heartbeat classification—have varying degrees of sensitivity to noise, heartbeat localization is the least sensitive, followed by reference point detection, while heartbeat classification is the most sensitive. Therefore, if the uncertainty of heartbeat localization for a given input signal is high, the uncertainties of reference point detection and heartbeat classification for that signal should also be higher. By summing the uncertainties of each task based on their respective noise sensitivities, the uncertainty corresponding to the heartbeat classification result is highest, followed by reference point detection, and then heartbeat localization, thus making the output uncertainty more consistent with prior knowledge. When applying this to training deep learning networks for multiple tasks, a higher uncertainty for noisy samples can reduce the backpropagation gradient, thereby reducing the interference of noise samples on training. Simultaneously, by summing the uncertainties, tasks with high noise sensitivity will have higher uncertainties, making them less affected by noise samples. In actual multi-task ECG analysis, the uncertainty output of the deep learning network for each task can be used to evaluate the reliability of the analysis results, such as by setting an uncertainty threshold to filter out reliable analysis results.

[0066] In one embodiment, the step of calculating the heartbeat positioning result based on the heartbeat positioning score and the heartbeat positioning uncertainty includes:

[0067] S404. Calculate the heartbeat location result using the following formula:

[0068]

[0069] Where, p loc For the heart rate localization results, s loc For heartbeat positioning score, u loc Let represent the uncertainty in heartbeat positioning, and let softmax be the normalized exponential function.

[0070] Since beat localization is least sensitive to noise, the uncertainty corresponding to the beat localization result calculation is set to the minimum, and the exponential coefficient is only the beat localization uncertainty. The softmax function is a normalized exponential function, which can be used to convert the output value into a probability distribution in the range [0, 1] and equal to 1.

[0071] In one embodiment, the step of calculating the reference point detection result based on the reference point detection score, the heartbeat positioning uncertainty, and the reference point detection uncertainty includes:

[0072] S405. Calculate the benchmark point detection results using the following formula:

[0073]

[0074] Where, p ael Based on the benchmark detection results, s del For the baseline detection score, u loc For the uncertainty of heartbeat positioning, u del The benchmark is used to determine the uncertainty, and the softmax function is a normalized exponential function.

[0075] Because reference point detection is more sensitive to noise than heartbeat localization, the exponential coefficient is set to the sum of the uncertainty of heartbeat localization and the uncertainty of reference point detection when calculating the uncertainty of reference point detection results. The softmax function is a normalized exponential function, which can convert the output value into a probability distribution in the range [0, 1] with a sum of 1.

[0076] In one embodiment, the step of calculating the heartbeat classification result based on the heartbeat classification score, the heartbeat positioning uncertainty, the reference point detection uncertainty, and the heartbeat classification uncertainty includes:

[0077] S406. Calculate the heart rate classification result using the following formula:

[0078]

[0079] Where, p cls For the heart-shot classification results, s cls Classify scores for heartbeats, u cls For the classification uncertainty of heartbeats, u loc For the uncertainty of heartbeat positioning, u del The benchmark is used to determine the uncertainty, and the softmax function is a normalized exponential function.

[0080] Since heartbeat classification is most sensitive to noise, the uncertainty corresponding to the calculation of the heartbeat classification result is set to the highest. The exponential coefficient is the sum of the heartbeat positioning uncertainty, the reference point detection uncertainty, and the heartbeat classification uncertainty. The softmax function is a normalized exponential function, which can convert the output value into a probability distribution ranging from [0, 1] to 1.

[0081] Reference Figure 2 This is a structural block diagram of an ECG multitasking analysis device according to an embodiment of this application. The device includes:

[0082] Acquisition module 100 is used to acquire electrocardiogram (ECG) signals;

[0083] The feature extraction module 200 is used to input the electrocardiogram (ECG) signal into the feature extraction network for feature extraction to obtain ECG signal features;

[0084] The multi-task network computing module 300 is used to input the ECG signal features into the deep learning networks corresponding to heartbeat localization, reference point detection, and heartbeat classification respectively for calculation, to obtain heartbeat localization score, reference point detection score, heartbeat classification score, heartbeat localization uncertainty, reference point detection uncertainty, and heartbeat classification uncertainty;

[0085] The electrocardiogram analysis module 400 is used to calculate, based on the heartbeat localization score, the reference point detection score, the heartbeat classification score, the heartbeat localization uncertainty, the reference point detection uncertainty, and the heartbeat classification uncertainty, to obtain the heartbeat localization result, the reference point detection result, and the heartbeat classification result.

[0086] In one embodiment, the above-mentioned ECG multitasking analysis device further includes:

[0087] The multi-task network calculation submodule is used to input the ECG signal features into the heartbeat localization network for calculation to obtain the heartbeat localization score and heartbeat localization uncertainty; input the ECG signal features into the reference point detection network for calculation to obtain the reference point detection score and reference point detection uncertainty; and input the ECG signal features into the heartbeat classification network for calculation to obtain the heartbeat classification score and heartbeat classification uncertainty.

[0088] In one embodiment, the above-mentioned ECG multitasking analysis device further includes:

[0089] The ECG analysis submodule is used to calculate the heartbeat localization result based on the heartbeat localization score and the heartbeat localization uncertainty; to calculate the reference point detection result based on the reference point detection score, the heartbeat localization uncertainty, and the reference point detection uncertainty; and to calculate the heartbeat classification result based on the heartbeat classification score, the heartbeat localization uncertainty, the reference point detection uncertainty, and the heartbeat classification uncertainty.

[0090] In one embodiment, the above-mentioned ECG multitasking analysis device further includes:

[0091] The first ECG analysis submodule is used to calculate the heartbeat localization result using the following formula:

[0092]

[0093] Where, p loc For the heart rate localization results, s loc For heartbeat positioning score, u loc Let represent the uncertainty in heartbeat positioning, and let softmax be the normalized exponential function.

[0094] In one embodiment, the above-mentioned ECG multitasking analysis device further includes:

[0095] The second ECG analysis submodule is used to calculate the baseline detection results using the following formula:

[0096]

[0097] Where, p del Based on the benchmark detection results, s del For the baseline detection score, u loc For the uncertainty of heartbeat positioning, u del The benchmark is used to determine the uncertainty, and the softmax function is a normalized exponential function.

[0098] In one embodiment, the above-mentioned ECG multitasking analysis device further includes:

[0099] The third ECG analysis submodule is used to calculate the heart rate classification results using the following formula:

[0100]

[0101] Where, p cls For the heart-shot classification results, s cls Classify scores for heartbeats, u cls For the classification uncertainty of heartbeats, u loc For the uncertainty of heartbeat positioning, u del The benchmark is used to determine the uncertainty, and the softmax function is a normalized exponential function.

[0102] Reference Figure 3 This application also provides a computer device, which may be a server, and its internal structure may be as follows: Figure 3 As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores data used in the ECG multi-task analysis method. The network interface allows communication with external terminals via a network connection. Furthermore, the computer device may also include input devices and a display screen. When the aforementioned computer program is executed by a processor, it employs a multi-task ECG analysis method, comprising the following steps: acquiring an ECG signal; inputting the ECG signal into a feature extraction network for feature extraction to obtain ECG signal features; inputting the ECG signal features into deep learning networks corresponding to heartbeat localization, reference point detection, and heartbeat classification for calculation to obtain heartbeat localization score, reference point detection score, heartbeat classification score, heartbeat localization uncertainty, reference point detection uncertainty, and heartbeat classification uncertainty; and calculating the heartbeat localization result, reference point detection result, and heartbeat classification result based on the heartbeat localization score, reference point detection score, heartbeat classification score, heartbeat localization uncertainty, reference point detection uncertainty, and heartbeat classification uncertainty.

[0103] Those skilled in the art will understand that Figure 3 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer equipment on which the present application is applied.

[0104] One embodiment of this application also provides a computer-readable storage medium storing a computer program. When executed by a processor, the computer program implements a multi-task ECG analysis method, including the following steps: acquiring an ECG signal; inputting the ECG signal into a feature extraction network for feature extraction to obtain ECG signal features; inputting the ECG signal features into deep learning networks corresponding to heartbeat localization, reference point detection, and heartbeat classification for calculation to obtain a heartbeat localization score, a reference point detection score, a heartbeat classification score, a heartbeat localization uncertainty, a reference point detection uncertainty, and a heartbeat classification uncertainty; calculating based on the heartbeat localization score, the reference point detection score, the heartbeat classification score, the heartbeat localization uncertainty, the reference point detection uncertainty, and the heartbeat classification uncertainty to obtain a heartbeat localization result, a reference point detection result, and a heartbeat classification result. It is understood that the computer-readable storage medium in this embodiment can be a volatile readable storage medium or a non-volatile readable storage medium.

[0105] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media provided in this application and in the embodiments may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual-speed SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAM bus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0106] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, apparatus, article, or method. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, apparatus, article, or method that includes that element.

[0107] The above description is only a preferred embodiment of this application and does not limit the patent scope of this application. Any equivalent structural or procedural changes made based on the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A method for multi-task analysis of electrocardiogram (ECG), characterized in that, The method includes: Acquire electrocardiogram (ECG) signals; The electrocardiogram (ECG) signal is input into a feature extraction network for feature extraction to obtain ECG signal features. The ECG signal features are input into the deep learning networks corresponding to heartbeat localization, reference point detection, and heartbeat classification, respectively, to calculate the heartbeat localization score, reference point detection score, heartbeat classification score, heartbeat localization uncertainty, reference point detection uncertainty, and heartbeat classification uncertainty. The heartbeat positioning score, the reference point detection score, the heartbeat classification score, the heartbeat positioning uncertainty, the reference point detection uncertainty, and the heartbeat classification uncertainty are calculated to obtain the heartbeat positioning result, the reference point detection result, and the heartbeat classification result. The step of calculating the heartbeat positioning result, the reference point detection result, and the heartbeat classification result based on the heartbeat positioning score, the reference point detection score, the heartbeat classification score, the heartbeat positioning uncertainty, the reference point detection uncertainty, and the heartbeat classification uncertainty includes: The heartbeat positioning result is obtained by calculating based on the heartbeat positioning score and the heartbeat positioning uncertainty; The reference point detection result is obtained by calculating the reference point detection score, the heartbeat positioning uncertainty, and the reference point detection uncertainty. The heartbeat classification result is obtained by calculating the heartbeat classification score, the heartbeat positioning uncertainty, the reference point detection uncertainty, and the heartbeat classification uncertainty. The step of calculating the heartbeat positioning result based on the heartbeat positioning score and the heartbeat positioning uncertainty includes: The heartbeat localization result is calculated using the following formula: in, For the heart-tracking location results, The score is determined by the heartbeat. For the uncertainty of heartbeat positioning, The function is a normalized exponential function; The step of calculating the reference point detection result based on the reference point detection score, the heartbeat positioning uncertainty, and the reference point detection uncertainty includes: The benchmark detection results are calculated using the following formula: in, Based on the benchmark detection results, The benchmark score is used to measure the score. For the uncertainty of heartbeat positioning, To detect uncertainty at the reference point, The function is a normalized exponential function; The step of calculating the heartbeat classification result based on the heartbeat classification score, the heartbeat positioning uncertainty, the reference point detection uncertainty, and the heartbeat classification uncertainty includes: The heart rate classification result is calculated using the following formula: in, Classification results for heartbeats Categorize the heartbeats into scores. For the classification uncertainty of heartbeats, For the uncertainty of heartbeat positioning, To detect uncertainty at the reference point, The function is a normalized exponential function.

2. The ECG multi-task analysis method according to claim 1, characterized in that, The step of inputting the ECG signal features into the deep learning networks corresponding to heartbeat localization, reference point detection, and heartbeat classification for calculation to obtain the heartbeat localization score, reference point detection score, heartbeat classification score, heartbeat localization uncertainty, reference point detection uncertainty, and heartbeat classification uncertainty includes: The ECG signal features are input into the heartbeat localization network for calculation to obtain the heartbeat localization score and heartbeat localization uncertainty. The ECG signal features are input into the reference point detection network for calculation to obtain the reference point detection score and the reference point detection uncertainty. The ECG signal features are input into a heartbeat classification network for calculation to obtain the heartbeat classification score and heartbeat classification uncertainty.

3. A multi-task electrocardiogram (ECG) analysis device, used to execute the multi-task ECG analysis method as described in claim 1, characterized in that, The device includes: The acquisition module is used to acquire electrocardiogram (ECG) signals; The feature extraction module is used to input the electrocardiogram (ECG) signal into the feature extraction network for feature extraction to obtain ECG signal features; The multi-task network computing module is used to input the ECG signal features into the deep learning networks corresponding to heartbeat localization, reference point detection, and heartbeat classification respectively for calculation, and obtain heartbeat localization score, reference point detection score, heartbeat classification score, heartbeat localization uncertainty, reference point detection uncertainty, and heartbeat classification uncertainty; The electrocardiogram (ECG) analysis module is used to calculate, based on the heartbeat localization score, the reference point detection score, the heartbeat classification score, the heartbeat localization uncertainty, the reference point detection uncertainty, and the heartbeat classification uncertainty, to obtain the heartbeat localization result, the reference point detection result, and the heartbeat classification result.

4. The ECG multitasking analysis device according to claim 3, characterized in that, include: The multi-task network calculation submodule is used to input the ECG signal features into the heartbeat localization network for calculation to obtain the heartbeat localization score and heartbeat localization uncertainty; input the ECG signal features into the reference point detection network for calculation to obtain the reference point detection score and reference point detection uncertainty; and input the ECG signal features into the heartbeat classification network for calculation to obtain the heartbeat classification score and heartbeat classification uncertainty.

5. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method of claim 1 or 2.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method of claim 1 or 2.

Images