Electrocardiosignal detection method and device, medium and electronic equipment
By using the target start and end position and third-order derivative position in pacing cardiac detection and performing feature enhancement processing on the initial electrocardiogram signal, the problem of low detection accuracy in the prior art is solved and the electrocardiogram signal is improved Sensitivity and accuracy of detection.
Patent Information
- Application Number
- CN202410903332.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-05
- Publication Date
- 2024-11-01
AI Technical Summary
The existing pacing cardiac detection technology is limited by the band segmentation method of electrocardiogram signals and interference noise, resulting in low detection accuracy, especially in the case of large deviations in training data.
By obtaining the initial ECG signal, determining the target start and end position and third-order derivative, positioning the pacing pulse, and performing feature enhancement processing on the initial ECG signal, inputting a pre-trained residual network model to improve the detection result.
The sensitivity and accuracy of the residual network model to detect ECG signals is improved, the detection difficulty is reduced, and the detection effect of ECG signals is improved.
Smart Images

Figure CN118873151A_ABST
Abstract
Description
Technical Field
[0001] The present disclosure belongs to the technical field of signal processing, and in particular relates to a method for detecting an electrocardiogram signal, an electrocardiogram signal detection device, a computer-readable storage medium, and an electronic device. Background Art
[0002] A paced heartbeat is an artificial heartbeat generated by a pacemaker through electrical impulse stimulation of the heart. In some cases, a patient may have both a natural heartbeat and a paced heartbeat. Therefore, accurate identification of paced heartbeats can help doctors distinguish between natural heart rhythm and the effects of pacemakers, help doctors make accurate diagnosis and treatment decisions to provide optimal heart rhythm management and heart health protection, and is of great significance for ensuring the normal operation of pacemakers, adjusting the paced heart rate, and monitoring the patient's heart status.
[0003] At present, the pacing heartbeat detection technology mainly detects the pacing heart rate by analyzing the electrocardiogram signal. However, the detection accuracy is completely limited by the segmentation method of each band of the electrocardiogram signal, and is also affected by other electrocardiogram interference or noise, resulting in inaccurate signal analysis. In addition, once encountering electrocardiogram data that deviates too much from the training data, the results are very unsatisfactory and the sensitivity is poor. Summary of the invention
[0004] In order to overcome the problems existing in the related art, the present disclosure provides a method for detecting an electrocardiogram signal, an electrocardiogram signal detection device, a computer-readable storage medium and an electronic device.
[0005] According to a first aspect of an embodiment of the present disclosure, a method for detecting an electrocardiogram signal is provided, the method comprising:
[0006] Acquire an initial electrocardiogram signal, and determine a target start and end position and a third-order derivative of the initial electrocardiogram signal according to the initial electrocardiogram signal;
[0007] Positioning a target pacing pulse according to the third-order derivative within the range where the target start and end positions are located;
[0008] Performing feature enhancement processing on the initial ECG signal according to the target pacing pulse to obtain a target ECG signal;
[0009] The target ECG signal is input into a pre-trained residual network model, so that the pre-trained residual network model outputs a detection result of the target ECG signal.
[0010] Optionally, determining the target start and end positions and the third-order derivative of the initial electrocardiogram signal according to the initial electrocardiogram signal includes:
[0011] Performing head-to-tail splicing processing on the initial electrocardiogram signal to obtain first sampling data, and determining the target start and end positions of the initial electrocardiogram signal according to the first sampling data;
[0012] The initial electrocardiogram signal is calculated to obtain a third-order derivative of the initial electrocardiogram signal.
[0013] Optionally, the step of performing head-to-tail concatenation processing on the initial electrocardiogram signal to obtain first sampling data includes:
[0014] Performing interception processing on the initial electrocardiogram signal to obtain second sampling data, and performing inversion processing on the second sampling data to obtain third sampling data;
[0015] The third sampling data and the initial electrocardiogram signal are spliced to obtain the first sampling data.
[0016] Optionally, determining the target start and end positions of the initial electrocardiogram signal according to the first sampling data includes:
[0017] Traversing the first sampling data to determine a peak position in the first sampling data, and determining a peak interval according to the peak position;
[0018] The initial start and end positions of the initial electrocardiogram signal are determined according to the peak position and the peak interval, and the initial start and end positions are filtered to obtain target start and end positions.
[0019] Optionally, locating the target pacing pulse according to the third-order derivative within the range where the target start and end positions are located includes:
[0020] Determining a target derivative in the third-order derivative within a range where the target start and end positions are located;
[0021] Determining a derivative mean of the third-order derivative according to the target derivative, and determining a target mean template according to the derivative mean;
[0022] When the target mean template and the derivative mean satisfy a preset condition, the target derivative corresponding to the derivative mean is determined as a target pacing pulse.
[0023] Optionally, determining the derivative mean of the third-order derivative according to the target derivative includes:
[0024] Acquire a plurality of third-order derivatives corresponding to the target derivative;
[0025] The target derivative and the plurality of third-order derivatives are calculated to obtain a derivative mean.
[0026] Optionally, determining a target mean template according to the derivative mean includes:
[0027] Determining a first mean template in the derivative mean, and determining a second mean template based on the derivative mean;
[0028] The first mean template and the second mean template are compared to obtain a comparison result, and a target mean template is determined in the first mean template and the second mean template according to the comparison result.
[0029] Optionally, performing feature enhancement processing on the initial electrocardiogram signal according to the target pacing pulse to obtain a target electrocardiogram signal includes:
[0030] performing a zeroing process on the third-order derivative according to the target pacing pulse to obtain a filtered third-order derivative;
[0031] Acquire a first amplitude corresponding to the initial electrocardiogram signal and a second amplitude corresponding to the filtered third-order derivative, and calculate the first amplitude and the second amplitude to obtain an enhancement coefficient;
[0032] The initial electrocardiogram signal is subjected to feature enhancement processing according to the enhancement coefficient and the filtered third-order derivative to obtain a target electrocardiogram signal.
[0033] According to a second aspect of an embodiment of the present disclosure, there is provided an electrocardiogram signal detection device, comprising:
[0034] A signal acquisition module, configured to acquire an initial ECG signal, and determine a target start and end position and a third-order derivative of the initial ECG signal according to the initial ECG signal;
[0035] A pulse positioning module, configured to position a target pacing pulse according to the third-order derivative within the range where the target start and end positions are located;
[0036] A feature enhancement module, configured to perform feature enhancement processing on the initial ECG signal according to the target pacing pulse to obtain a target ECG signal;
[0037] The model detection module is configured to input the target ECG signal into a pre-trained residual network model so that the pre-trained residual network model outputs a detection result of the target ECG signal.
[0038] According to a third aspect of an embodiment of the present disclosure, a computer-readable storage medium is provided, on which computer program instructions are stored. When the program instructions are executed by a processor, the steps of any one of the electrocardiogram signal detection methods provided in the first aspect of the present disclosure are implemented.
[0039] According to a fourth aspect of an embodiment of the present disclosure, there is provided an electronic device, including:
[0040] processor;
[0041] a memory for storing processor-executable instructions;
[0042] The processor is configured to: execute the executable instructions to implement the steps of any one of the electrocardiogram signal detection methods provided in the first aspect of the present disclosure.
[0043] The technical solution provided by the embodiments of the present disclosure may have the following beneficial effects:
[0044] In the method and apparatus provided in the exemplary embodiments of the present disclosure, the target start and end positions and the third-order derivatives can be used to locate the target pacing pulse beat by beat, providing a data basis and theoretical support for the residual network model to output the detection result of the target ECG signal, improving the sensitivity of the residual network model in detecting ECG signals, and optimizing the detection effect of the residual network model. Furthermore, the initial ECG signal is processed with feature enhancement using the located target pacing pulse, providing input data for the residual network model, enhancing the features of the ECG signal, reducing the difficulty of detection and classification of the residual network model, and improving the detection accuracy and effectiveness of the ECG signal.
[0045] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS
[0046] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.
[0047] Figure 1 A schematic diagram of a process of detecting an electrocardiogram signal in an exemplary embodiment of the present disclosure is schematically shown;
[0048] Figure 2 A schematic flowchart of a method for determining a target start and end position and a third-order derivative in an exemplary embodiment of the present disclosure is schematically shown;
[0049] Figure 3 The flowchart of the method for performing head-to-tail splicing processing on the initial electrocardiogram signal in the exemplary embodiment of the present disclosure is schematically shown;
[0050] Figure 4 A schematic flow chart of a method for determining the start and end positions of a target according to first sampling data in an exemplary embodiment of the present disclosure is schematically shown;
[0051] Figure 5 A schematic flowchart of a method for locating a target pacing pulse in an exemplary embodiment of the present disclosure is schematically shown;
[0052] Figure 6 A schematic diagram of a flow chart of a method for determining a derivative mean value according to a target derivative in an exemplary embodiment of the present disclosure is schematically shown;
[0053] Figure 7 A schematic diagram of a process for determining a target mean template according to a derivative mean in an exemplary embodiment of the present disclosure is shown;
[0054] Figure 8 A schematic flow chart of a method for performing feature enhancement processing on an initial electrocardiogram signal in an exemplary embodiment of the present disclosure is schematically shown;
[0055] Fig. 9 A schematic diagram schematically illustrates a framework of a method for detecting an electrocardiogram signal in an application scenario in an exemplary embodiment of the present disclosure;
[0056] Fig.10 A schematic diagram schematically shows the structure of an electrocardiogram signal detection device in an exemplary embodiment of the present disclosure;
[0057] Fig.11 A schematic diagram schematically shows the structure of another electrocardiogram signal detection device in an exemplary embodiment of the present disclosure;
[0058] Fig.12 The following is a schematic diagram showing the structure of another electrocardiogram signal detection device in an exemplary embodiment of the present disclosure. DETAILED DESCRIPTION
[0059] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.
[0060] It should be noted that all actions of acquiring signals, information or data in the present disclosure are carried out in compliance with the relevant data protection laws and policies of the country where the device is located and with the authorization given by the owner of the corresponding device.
[0061] A paced heartbeat is an artificial heartbeat generated by a pacemaker through electrical impulse stimulation of the heart. In some cases, a patient may have both a natural heartbeat and a paced heartbeat. Therefore, accurate identification of paced heartbeats can help doctors distinguish between natural heart rhythm and the effects of pacemakers, help doctors make accurate diagnosis and treatment decisions to provide optimal heart rhythm management and heart health protection, and is of great significance for ensuring the normal operation of pacemakers, adjusting the paced heart rate, and monitoring the patient's heart status.
[0062] With the development of aging, pacemakers will be accepted by more and more elderly people, and the probability of pacemaker heartbeats will greatly increase. The intelligent detection technology of pacemaker heartbeats will also have great application potential in hospitals, physical examination centers and health care centers.
[0063] At present, pacing heart beat detection technology mainly detects pacing heart rate by analyzing electrocardiogram signals. The specific technologies are divided into ECG signal rhythm analysis, pacing signal detection and deep learning methods.
[0064] Among them, ECG signal rhythm analysis means that ECG signals can display the rhythm of the heart, including sinus rhythm, pacemaker-paced atrial rhythm and ventricular rhythm, etc. By analyzing the characteristics of the R wave (the first upward QRS wave, the QRS wave group is the ventricular depolarization wave in the ECG) and P wave (waveform of atrial depolarization) on the ECG, the pacing of the heart and the working mode of the pacemaker can be determined. However, the detection accuracy is completely limited by the segmentation method of each band of the ECG signal.
[0065] Pacing signal detection refers to the determination of the pacemaker output by analyzing the pacing signal on the electrocardiogram. However, pacing signal detection may be affected by other electrocardiogram interference or noise, resulting in inaccurate signal analysis.
[0066] Deep learning methods can automatically extract features from electrocardiograms, avoiding the tediousness and omissions of manual extraction of electrocardiogram features. With the support of large amounts of data and long-term training, the effect is good. However, once encountering electrocardiogram data that deviates too much from the training data, the results are very unsatisfactory and the sensitivity is poor.
[0067] In view of the problems existing in the related art, the present invention provides a method for detecting an electrocardiogram signal. Figure 1 is a flow chart of a method for detecting an electrocardiogram signal according to an exemplary embodiment. Figure 1 As shown, the method may at least include the following steps:
[0068] Step S110: Acquire an initial ECG signal, and determine the target start and end positions and the third-order derivative of the initial ECG signal according to the initial ECG signal.
[0069] Step S120. Locate the target pacing pulse according to the third-order derivative within the range of the target start and end positions.
[0070] Step S130: Perform feature enhancement processing on the initial ECG signal according to the target pacing pulse to obtain a target ECG signal.
[0071] Step S140: Input the target ECG signal into the pre-trained residual network model, so that the pre-trained residual network model outputs the detection result of the target ECG signal.
[0072] In the exemplary embodiments of the present disclosure, the target start and end positions and the third-order derivatives can be used to locate the target pacing pulse beat by beat, providing a data basis and theoretical support for the residual network model to output the detection result of the target ECG signal, improving the sensitivity of the residual network model in detecting ECG signals, and optimizing the detection effect of the residual network model. Furthermore, the initial ECG signal is processed with feature enhancement using the located target pacing pulse, providing input data for the residual network model, enhancing the features of the ECG signal, reducing the difficulty of detection and classification of the residual network model, and improving the detection accuracy and effectiveness of the ECG signal.
[0073] The following is a detailed description of each step of the method for detecting an electrocardiogram signal.
[0074] In step S110, an initial electrocardiogram signal is acquired, and target start and end positions and a third-order derivative of the initial electrocardiogram signal are determined according to the initial electrocardiogram signal.
[0075] In an exemplary embodiment of the present disclosure, the initial ECG signal may be a 12-lead 10-second ECG signal with a sampling frequency of 500 Hz (Hertz), denoted as S1.
[0076] In an alternative embodiment, Figure 2 A flow chart of a method for determining the target start and end positions and the third-order derivative is shown, as shown in FIG. Figure 2 As shown, the method may at least include the following steps: In step S210, the initial ECG signal is processed by head-to-tail concatenation to obtain first sampling data, and the target start and end positions of the initial ECG signal are determined according to the first sampling data.
[0077] In an alternative embodiment, Figure 3 FIG. 4 is a flow chart showing a method for processing the initial ECG signal by connecting the initial ECG signal to the end. Figure 3 As shown, the method may at least include the following steps: in step S310, the initial electrocardiogram signal is intercepted to obtain second sampling data, and the second sampling data is inverted to obtain third sampling data.
[0078] For the initial electrocardiogram signal, the first 2000 sampling point data and the last 2000 sampling point data can be intercepted as the second sampling data.
[0079] Furthermore, the first 2000 sampling point data in the second sampling data are reversed, which is recorded as S1. B2000’; The last 2000 sampling points in the second sampling data are reversed, denoted as S1 E2000’ At this time, the inverted S1 B2000’ and S1 E2000’ collectively serve as the third sampling data.
[0080] Among them, the reversal processing method can be to reverse the order of the first 2000 sampling point data intercepted in the second sampling data, and reverse the order of the last 2000 sampling point data intercepted in the second sampling data. Other reversal processing methods can also be selected according to actual conditions. This exemplary embodiment does not make special limitations on this.
[0081] In step S320, the third sampling data and the initial electrocardiogram signal are concatenated to obtain the first sampling data.
[0082] After obtaining the third sampling data, you can follow S1 B2000’ , S1, S1 E2000’ The third sampling data and the initial electrocardiogram signal are spliced into a data with a length of 9000 sampling points as the first sampling data, and the first sampling data is recorded as S2.
[0083] In an alternative embodiment, Figure 4 A flow chart of a method for determining the start and end positions of a target according to first sampling data is shown, as shown in FIG. Figure 4 As shown, the method may at least include the following steps: In step S410, the first sampling data is traversed to determine the peak position in the first sampling data, and the peak interval is determined according to the peak position.
[0084] After the first sampling data is obtained, the wfdb.processing.gqrs_detect function can be used to traverse leads I and II to capture the R peak (representing the contraction of the heart) position of the first sampling data S2 as the peak position.
[0085] Furthermore, the peak interval is calculated using formula (1):
[0086] RR_interval (i) =R_index (i+1) -R_index (i) (1)
[0087] Among them, RR_interval (i) is the i-th RR (the time between two R waves on the electrocardiogram) interval, R_index (i) is the position of the ith R peak, R_index (i+1) is the i+1th R peak position.
[0088] In step S420, the initial start and end positions of the initial ECG signal are determined according to the peak position and the peak interval, and the initial start and end positions are filtered to obtain the target start and end positions.
[0089] Except for the first and last beats, the starting and ending points of each remaining ECG signal are calculated using formula (2) as the initial starting and ending positions:
[0090]
[0091] Among them, Beg_index (i) is the starting position of the ith heartbeat, End_index (i) is the end position of the ith heartbeat.
[0092] Since the first sampling data is obtained by adding 2000 inverted sampling point data at the beginning and the end, the heartbeats with a starting position less than 2000 and the heartbeats with an ending position greater than 7000 can be filtered out to obtain the target starting and ending positions of the initial ECG signal.
[0093] In step S220, the initial ECG signal is calculated to obtain the third-order derivative of the initial ECG signal.
[0094] On the other hand, the third-order derivative of the 10-second initial ECG signal is calculated and recorded as Dif_ecg 3 .
[0095] In step S120, the target pacing pulse is located according to the third-order derivative within the range of the target start and end positions.
[0096] In an exemplary embodiment of the present disclosure, after determining the target start and end positions and the third-order derivative, the target pacing pulse can be further positioned.
[0097] In an alternative embodiment, Figure 5 A flow chart of a method for locating a target pacing pulse is shown, Figure 5 As shown, the method may at least include the following steps: In step S510, within the range where the target start and end positions are located, a target derivative is determined in the third-order derivative.
[0098] Within the range represented by the target start and end positions of the i-th ECG data, find the maximum value in the third-order derivative as the target derivative, and record the corresponding position as P_d 3 i Similarly, by traversing each ECG signal in turn, we can get a list of the peak positions of the third-order derivatives of all heart beats within the 10-second initial ECG signal, denoted as P_d 3 .
[0099] In step S520, a derivative mean of the third-order derivative is determined according to the target derivative, and a target mean template is determined according to the derivative mean.
[0100] In an alternative embodiment, Figure 6 A flow chart of a method for determining a derivative mean value according to a target derivative is shown, as Figure 6 As shown, the method may at least include the following steps: In step S610, a plurality of third-order derivatives corresponding to the target derivatives are obtained.
[0101] After obtaining the target derivative, you can take the target derivative P_d 3 i The first three sampling points to P_d 3 i The data of the last four sampling points are recorded as PR_d 3 i , which is the data near the peak value of the third-order derivative of the i-th beat ECG signal.
[0102] In step S620, the target derivative and a plurality of third-order derivatives are calculated to obtain a derivative mean.
[0103] Furthermore, the target derivative and multiple third-order derivatives are averaged and the average is taken as the derivative mean, which is recorded as
[0104] Similarly, traverse P_d in sequence 3 For each peak in , we can get the mean list near the peak of the third-order derivative of all heart beats in the 10-second initial ECG signal, recorded as
[0105] In an alternative embodiment, Figure 7 A flow chart of a method for determining a target mean template based on a derivative mean is shown, as Figure 7 As shown, the method may at least include the following steps: in step S710, a first mean template is determined in the derivative mean, and a second mean template is determined according to the derivative mean.
[0106] In the list of derivative means In the above example, the third and fourth quartiles of the data in the list are determined as the first mean template near the peak of the third-order derivative, denoted as
[0107] Furthermore, in the list of derivative means In the example, 0.2 times the maximum value of the data in the list can also be used as the second mean template.
[0108] In step S720, the first mean template and the second mean template are compared to obtain a comparison result, and a target mean template is determined from the first mean template and the second mean template according to the comparison result.
[0109] After determining the first mean template and the second mean template, if the first mean template of the three quartiles is less than 0.2 times the second mean template determined by the maximum value of the data in the list, the second mean template is set as the mean template near the peak of the third-order derivative, that is, the target mean template, otherwise the first mean template is determined as the target mean template. The target mean template can be adaptively adjusted with each 10-second initial ECG data.
[0110] In step S530, when the target mean template and the derivative mean meet preset conditions, the target derivative corresponding to the derivative mean is determined as the target pacing pulse.
[0111] Among them, the preset conditions satisfied by the target mean template and the derivative mean are shown in formula (3):
[0112]
[0113] When formula (3) is satisfied, P_d can be determined 3 i is the pseudo pacing pulse, that is, the target pacing pulse, denoted as PP_d 3 i . Similarly, traverse P_d in turn 3 Pseudo pacing pulse list PP_d is available 3 .
[0114] In step S130, feature enhancement processing is performed on the initial ECG signal according to the target pacing pulse to obtain a target ECG signal.
[0115] In an exemplary embodiment of the present disclosure, after locating the target pacing pulse, the target pacing pulse may be used to perform feature enhancement processing on the initial ECG signal.
[0116] In an alternative embodiment, Figure 8 A flow chart of a method for performing feature enhancement processing on an initial ECG signal is shown, Figure 8 As shown, the method may at least include the following steps: In step S810, the third-order derivative is zeroed according to the target pacing pulse to obtain a filtered third-order derivative.
[0117] According to the position of the target pacing pulse PP_d 3 i , retain PP_d 3 i The first three sampling points to PP_d 3i The data of the last four sampling points are used to set the data of the third-order derivative sampling points of the ECG signal at the non-target pacing pulse positions to zero, and the filtered third-order derivative data of the ECG signal Dif_ecg is obtained. 3 filtered .
[0118] In step S820, a first amplitude corresponding to the initial electrocardiogram signal and a second amplitude corresponding to the filtered third-order derivative are obtained, and the first amplitude and the second amplitude are calculated to obtain an enhancement coefficient.
[0119] Furthermore, the maximum amplitude of the initial ECG signal is calculated as the first amplitude, denoted as F ecg , and calculate the third-order derivative Dif_ecg of 10 seconds of ECG data 3 The maximum amplitude is taken as the second amplitude, denoted as Fd 3 .
[0120] The enhancement coefficient Ad_value is calculated according to formula (4):
[0121]
[0122] In step S830, feature enhancement processing is performed on the initial ECG signal according to the enhancement coefficient and the filtered third-order derivative to obtain a target ECG signal.
[0123] According to formula (5), the target ECG signal Fus_ecg after feature fusion is calculated:
[0124] Fus_ecg=Ad_value*Dif_ecg 3 filtered +S1(5)
[0125] In step S140, the target ECG signal is input into a pre-trained residual network model, so that the pre-trained residual network model outputs a detection result of the target ECG signal.
[0126] In an exemplary embodiment of the present disclosure, the training data set used when training the pre-trained residual network model can be A large scale 12-lead electrocardio gram databasefor arrhythmia study data set on the PhysioNet website and PTB-XL, a large publicly available electrocardiography dataset.
[0127] It is worth noting that since the main focus is on the presence or absence of paced heartbeats, other diseases are not taken into consideration, so the data labels are only divided into two types according to needs: paced and non-paced. The 10-second standard ECG signal Fus_ecg after feature fusion is encapsulated into a matrix of 1*1*12*5000 for use by the deep learning algorithm.
[0128] Among them, the pre-trained residual network model uses the torchvision.models.resnet50 function to call the resnet50 deep learning network and feeds in a packaged 1*1*12*5000 matrix for training.
[0129] During model training, 30% of the data is randomly divided into the test set and 70% is the training set; the batch-size is 8; the loss function is CrossEntropyLoss (cross entropy loss function), the optimizer is Adam (a gradient-based optimization algorithm), the learning rate is 0.0005, and the epoch (cycle) is 400.
[0130] During model inference, the target ECG signal and the trained weights are loaded into the model to obtain the inference result, which indicates whether the 10-second initial ECG signal contains a pacemaker beat.
[0131] It is worth noting that when the deep learning classification result shows that the 10-second initial ECG signal contains a paced heartbeat, the located target paced pulse list PP_d 3 , which is a list of pacing pulse positions located beat by beat.
[0132] The following is a detailed description of the method for detecting an electrocardiogram signal in an embodiment of the present disclosure in conjunction with an application scenario.
[0133] Fig. 9 A schematic diagram of a detection method for an electrocardiogram signal in an application scenario is shown, Fig. 9 As shown in the figure, the new framework for beat-by-beat detection of pacing heartbeats based on deep learning and the third-order derivative of ECG signals consists of four parts: ECG signal preprocessing method, beat-by-beat positioning of pseudo pacing pulses, ECG signal feature enhancement and pacing heart rate classification.
[0134] First, the ECG signal is preprocessed, which is divided into three steps: concatenating the beginning and the end of the ECG signal, segmenting the RR interval of the entire data segment, and calculating the third-order derivative of the ECG data. Secondly, the stage of locating the pseudo-pacing pulse beat by beat is divided into four parts: locating the peak of the third-order derivative of the ECG signal, calculating the mean near the derivative peak, adaptively setting the mean template near the derivative peak, and locating the pseudo-pacing pulse. Then, the stage of feature enhancement of the ECG signal is divided into three parts: filtering non-pseudo-pacing pulse data, data standardization, and data feature fusion. Next, the stage of pacing heart rate classification is divided into two parts: making a data set and building a ResNet (deep residual network). Finally, the results of beat-by-beat detection and positioning of the pacemaker heart beat can be obtained.
[0135] Specifically, in the ECG signal preprocessing stage, the ECG signal is spliced head to tail to complete the first and last heart beats; and the wfdb (open source software package) toolkit is used to obtain all R peaks of a 10-second ECG data segment, and then the RR interval of the 10-second ECG full data segment is calculated to obtain the approximate starting and ending points of each ECG signal; finally, the third-order derivative of the ECG data is calculated.
[0136] In step S901, the ECG signal is spliced end to end.
[0137] The input data are all 12-lead 10-second ECG signals with a sampling frequency of 500Hz, denoted as S1. The first 2000 sampling points are intercepted and the data is inverted, denoted as S1 B2000’ . The last 2000 sampling points are intercepted and the data is reversed, recorded as S1 E2000’ Finally, according to S1 B2000’ , S1, S1 E2000’ These three segments of data are sequentially concatenated into a data set with a length of 9000 sampling points, which is recorded as S2.
[0138] In step S902, the RR interval of the entire data segment is segmented. The R peak position of the S2 data can be captured by traversing leads I and II through the wfdb.processing.gqrs_detect function, and the RR interval is calculated using formula (1). Except for the first and last beats, the start and end points of each remaining ECG signal are calculated using formula (2). Finally, the heartbeats with a start position less than 2000 and the heartbeats with an end position greater than 7000 are filtered out.
[0139] In step S903, the third-order derivative of the ECG data is calculated. The third-order derivative of the 10-second ECG data is calculated and recorded as Dif_ecg 3 .
[0140] At this point, after preprocessing the 10-second ECG data, all RR intervals, the start and end data segments of each heart beat, and the third-order derivative data of the ECG signal were obtained.
[0141] In the stage of locating pseudo-pacing pulses beat by beat, the peak value of the third-order derivative of the ECG signal of each heartbeat is used to calculate the mean value near the peak value of the third-order derivative; then the mean value near the peak value of the third-order derivative of the ECG signal of each beat is traversed, and the mean template near the peak value of the third-order derivative is adaptively set according to the three quartiles; finally, the mean template is compared with the third-order derivative of the ECG signal of each beat to achieve the purpose of locating pseudo-pacing pulses beat by beat.
[0142] In step S904, the peak of the third-order derivative of the ECG signal is located.
[0143] In the start and end range of the i-th ECG data, find the maximum value of the third-order derivative, and its position is recorded as P_d 3 i By traversing each ECG signal in turn, we can obtain a list of the peak positions of the third-order derivatives of all heartbeats within a 10-second ECG signal, denoted as P_d 3 .
[0144] In step S905 , the mean value near the derivative peak is calculated.
[0145] Take the peak value of the third-order derivative P_d 3 i The first three sampling points to P_d 3 i The data of the next four sampling points are recorded as PR_d 3 i , which is the data near the peak value of the third-order derivative of the i-th ECG signal, and finally the average is calculated and recorded as
[0146] In step S906, a mean template of the derivative peak is adaptively set.
[0147] According to the method of STEP2, traverse P_d in sequence 3 For each peak in , we can get the mean list near the peak of the third-order derivative of all heart beats in the 10-second ECG signal, recorded as Take the third and fourth quartiles of the data in the list as the mean template near the peak of the third-order derivative, denoted as If the third and fourth quartiles are less than 0.2 times the maximum value of the data in the list, the maximum value is set as the mean template near the peak value of the third-order derivative. The mean template can be adaptively adjusted with each 10-second ECG data.
[0148] In step S907, the pseudo pacing pulse is located.
[0149] If formula (3) is satisfied, then P_d 3 i is a pseudo pacing pulse, denoted as PP_d 3 i . ; Traverse P_d in sequence3 Pseudo pacing pulse list PP_d is available 3 .
[0150] In the feature enhancement stage, the pacing pulses in some paced heart beats are too small, which increases the difficulty of classification and reduces the classification accuracy. To address this problem, the method of fusing the third-order derivative of the ECG signal with the ECG signal is used to enhance the signal features to reduce the difficulty of the classification task.
[0151] Firstly, the non-pseudo pacing pulse data is filtered according to the located position of the pseudo pacing pulse; the third-order derivative of the filtered ECG signal is normalized according to the maximum amplitude of the original ECG signal; finally, it is superimposed with the ECG signal for feature fusion.
[0152] In step S908, non-false pacing pulse data is filtered.
[0153] According to the located pseudo pacing pulse position PP_d 3 i , retain PP_d 3 i The first three sampling points to PP_d 3 i The data of the last four sampling points are used to set the data of the third-order derivative sampling points of the ECG signal at the non-pseudo-pacing pulse positions to zero, and the filtered third-order derivative data of the ECG signal Dif_ecg is obtained. 3 filtered .
[0154] In step S909, the data is normalized.
[0155] Specifically, calculate the maximum amplitude of the original ECG signal, denoted as F ecg ; And, calculate the third-order derivative Dif_ecg of 10 seconds of ECG data 3 The maximum amplitude is denoted as Fd 3 Furthermore, the data adjustment coefficient Ad_value can be calculated according to formula (4).
[0156] In step S910, data-feature fusion.
[0157] The ECG data Fus_ecg after feature fusion is calculated according to formula (5).
[0158] In the pacing heart rate classification stage, the A large scale 12-lead electrocardiogram database for arrhythmia study dataset on the PhysioNet website and PTB-XL, a large publicly available electrocardiography dataset, were used to train the Resnet50 network under the Pyhorch deep learning framework for pacing heart rate classification.
[0159] In step S911, a data set is created.
[0160] Specifically, the training datasets used are A large scale 12-lead electrocardiogram database for arrhythmia study dataset and PTB-XL, a large publicly available electrocardiography dataset on the PhysioNet website.
[0161] It is worth noting that, considering that the main focus is on the presence or absence of paced heartbeats, other types of diseases are not taken into consideration, and their data labels are only divided into two types: paced and non-paced according to needs.
[0162] The 10-second ECG data Fus_ecg after feature fusion is encapsulated into a matrix of 1*1*12*5000 for use by the deep learning algorithm.
[0163] In step S912, a ResNet network is built.
[0164] Use the torchvision.models.resnet50 function to call the resnet50 deep learning network and feed in a packaged 1*1*12*5000 matrix for training.
[0165] During model training, 30% of the data is randomly divided into the test set and 70% is the training set; the batch-size is 8; the loss function is CrossEntropyLoss, the optimizer is Adam, the learning rate is 0.0005, and the epoch is 400.
[0166] During model inference, the data is processed according to the above steps, and the data and the trained weights are loaded into the model to obtain the inference result, which indicates whether the 10-second ECG data contains a paced heartbeat.
[0167] In step S913, the pacemaker heartbeat is detected and located beat by beat.
[0168] If the deep learning classification result shows that the 10-second ECG data contains a paced heartbeat, then the pseudo paced pulse list PP_d3 is a list of paced pulse positions located beat by beat.
[0169] In this application scenario, based on the method of deep learning and the third-order derivative of ECG signals, intelligent detection of pacemaker heart beat by beat is realized. Generally, it can be deployed on Windows and Linux operating systems, and does not require a graphics card, only the CPU is needed to complete model training and reasoning, thereby realizing a low-cost, easy-to-deploy, lightweight detection system for pacemaker heart beat by beat. The evaluation indicators of its test set are shown in Table 1, which shows that the performance of the invention has reached a practical level.
[0170]
[0171] Table 1
[0172] In the exemplary embodiments of the present disclosure, the target start and end positions and the third-order derivatives can be used to locate the target pacing pulse beat by beat, providing a data basis and theoretical support for the residual network model to output the detection result of the target ECG signal, improving the sensitivity of the residual network model in detecting ECG signals, and optimizing the detection effect of the residual network model. Furthermore, the initial ECG signal is processed with feature enhancement using the located target pacing pulse, providing input data for the residual network model, enhancing the features of the ECG signal, reducing the difficulty of detection and classification of the residual network model, and improving the detection accuracy and effectiveness of the ECG signal.
[0173] In addition, in an exemplary embodiment of the present disclosure, a device for detecting an electrocardiogram signal is also provided. Fig.10 FIG. 4 shows a schematic diagram of the structure of a detection device for an electrocardiogram signal. Fig.10 As shown, the electrocardiogram signal detection device 1000 may include: a signal acquisition module 1010, a pulse location module 1020, a feature enhancement module 1030 and a model detection module 1040. Among them:
[0174] The signal acquisition module 1010 is configured to acquire an initial ECG signal, and determine a target start and end position and a third-order derivative of the initial ECG signal according to the initial ECG signal;
[0175] The pulse positioning module 1020 is configured to position the target pacing pulse according to the third-order derivative within the range where the target start and end positions are located;
[0176] The feature enhancement module 1030 is configured to perform feature enhancement processing on the initial ECG signal according to the target pacing pulse to obtain a target ECG signal;
[0177] The model detection module 1040 is configured to input the target ECG signal into a pre-trained residual network model so that the pre-trained residual network model outputs a detection result of the target ECG signal.
[0178] In some embodiments of the present disclosure, the signal acquisition module 1010 is configured to:
[0179] Performing head-to-tail splicing processing on the initial electrocardiogram signal to obtain first sampling data, and determining the target start and end positions of the initial electrocardiogram signal according to the first sampling data;
[0180] The initial electrocardiogram signal is calculated to obtain a third-order derivative of the initial electrocardiogram signal.
[0181] In some embodiments of the present disclosure, the signal acquisition module 1010 is configured to:
[0182] Performing interception processing on the initial electrocardiogram signal to obtain second sampling data, and performing inversion processing on the second sampling data to obtain third sampling data;
[0183] The third sampling data and the initial electrocardiogram signal are spliced to obtain the first sampling data.
[0184] In some embodiments of the present disclosure, the signal acquisition module 1010 is configured to:
[0185] Traversing the first sampling data to determine a peak position in the first sampling data, and determining a peak interval according to the peak position;
[0186] The initial start and end positions of the initial electrocardiogram signal are determined according to the peak position and the peak interval, and the initial start and end positions are filtered to obtain target start and end positions.
[0187] In some embodiments of the present disclosure, the pulse positioning module 1020 is configured to:
[0188] Determining a target derivative in the third-order derivative within a range where the target start and end positions are located;
[0189] Determining a derivative mean of the third-order derivative according to the target derivative, and determining a target mean template according to the derivative mean;
[0190] When the target mean template and the derivative mean satisfy a preset condition, the target derivative corresponding to the derivative mean is determined as a target pacing pulse.
[0191] In some embodiments of the present disclosure, the pulse positioning module 1020 is configured to:
[0192] Acquire a plurality of third-order derivatives corresponding to the target derivative;
[0193] The target derivative and the plurality of third-order derivatives are calculated to obtain a derivative mean.
[0194] In some embodiments of the present disclosure, the pulse positioning module 1020 is configured to:
[0195] Determining a first mean template in the derivative mean, and determining a second mean template based on the derivative mean;
[0196] The first mean template and the second mean template are compared to obtain a comparison result, and a target mean template is determined in the first mean template and the second mean template according to the comparison result.
[0197] In some embodiments of the present disclosure, the feature enhancement module 1030 is configured to:
[0198] performing a zeroing process on the third-order derivative according to the target pacing pulse to obtain a filtered third-order derivative;
[0199] Acquire a first amplitude corresponding to the initial electrocardiogram signal and a second amplitude corresponding to the filtered third-order derivative, and calculate the first amplitude and the second amplitude to obtain an enhancement coefficient;
[0200] The initial electrocardiogram signal is subjected to feature enhancement processing according to the enhancement coefficient and the filtered third-order derivative to obtain a target electrocardiogram signal.
[0201] Regarding the device in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.
[0202] The present disclosure also provides a computer-readable storage medium having computer program instructions stored thereon, and when the program instructions are executed by a processor, the steps of the electrocardiogram signal detection method provided by the present disclosure are implemented.
[0203] Fig.11 1 is a block diagram of another device 1100 for detecting an electrocardiogram signal according to an exemplary embodiment. For example, the device 1100 may be a mobile phone, a computer, a digital broadcast terminal, a messaging device, a game console, a tablet device, a medical device, a fitness device, a personal digital assistant, etc.
[0204] Reference Fig.11, the device 1100 may include one or more of the following components: a processing component 1102 , a memory 1104 , a power component 1106 , a multimedia component 1108 , an audio component 1110 , an input / output interface 1112 , a sensor component 1114 , and a communication component 1116 .
[0205] The processing component 1102 generally controls the overall operation of the device 1100, such as operations associated with display, phone calls, data communications, camera operations, and recording operations. The processing component 1102 may include one or more processors 1120 to execute instructions to complete all or part of the steps of the above method. In addition, the processing component 1102 may include one or more modules to facilitate the interaction between the processing component 1102 and other components. For example, the processing component 1102 may include a multimedia module to facilitate the interaction between the multimedia component 1108 and the processing component 1102.
[0206] The memory 1104 is configured to store various types of data to support the operation of the device 1100. Examples of such data include instructions for any application or method operating on the device 1100, contact data, phone book data, messages, pictures, videos, etc. The memory 1104 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.
[0207] The power supply component 1106 provides power to the various components of the device 1100. The power supply component 1106 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power for the device 1100.
[0208] The multimedia component 1108 includes a screen that provides an output interface between the device 1100 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor may not only sense the boundaries of the touch or slide action, but also detect the duration and pressure associated with the touch or slide operation. In some embodiments, the multimedia component 1108 includes a front camera and / or a rear camera. When the device 1100 is in an operating mode, such as a shooting mode or a video mode, the front camera and / or the rear camera may receive external multimedia data. Each front camera and rear camera may be a fixed optical lens system or have a focal length and optical zoom capability.
[0209] The audio component 1110 is configured to output and / or input audio signals. For example, the audio component 1110 includes a microphone (MIC), and when the device 1100 is in an operating mode, such as a call mode, a recording mode, and a speech recognition mode, the microphone is configured to receive an external audio signal. The received audio signal can be further stored in the memory 1104 or sent via the communication component 1116. In some embodiments, the audio component 1110 also includes a speaker for outputting audio signals.
[0210] The input / output interface 1112 provides an interface between the processing component 1102 and the peripheral interface modules, which may be keyboards, click wheels, buttons, etc. These buttons may include but are not limited to: a home button, a volume button, a start button, and a lock button.
[0211] The sensor assembly 1114 includes one or more sensors for providing various aspects of the status assessment of the device 1100. For example, the sensor assembly 1114 can detect the open / closed state of the device 1100, the relative positioning of components, such as the display and keypad of the device 1100, the sensor assembly 1114 can also detect the position change of the device 1100 or a component of the device 1100, the presence or absence of user contact with the device 1100, the orientation or acceleration / deceleration of the device 1100, and the temperature change of the device 1100. The sensor assembly 1114 can include a proximity sensor configured to detect the presence of a nearby object without any physical contact. The sensor assembly 1114 can also include an optical sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly 1114 can also include an accelerometer, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.
[0212] The communication component 1116 is configured to facilitate wired or wireless communication between the device 1100 and other devices. The device 1100 can access a wireless network based on a communication standard, such as WiFi, 2G or 3G, or a combination thereof. In an exemplary embodiment, the communication component 1116 receives a broadcast signal or broadcast-related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 1116 also includes a near field communication (NFC) module to facilitate short-range communication. For example, the NFC module can be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.
[0213] In an exemplary embodiment, the device 1100 can be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components to perform the above methods.
[0214] In an exemplary embodiment, a non-transitory computer-readable storage medium including instructions is also provided, such as a memory 1104 including instructions, and the instructions can be executed by the processor 1120 of the device 1100 to perform the above method. For example, the non-transitory computer-readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc.
[0215] In addition to being an independent electronic device, the above-mentioned device can also be a part of an independent electronic device. For example, in one embodiment, the device can be an integrated circuit (IC) or a chip, wherein the integrated circuit can be an IC or a collection of multiple ICs; the chip can include but is not limited to the following types: GPU (Graphics Processing Unit), CPU (Central Processing Unit), FPGA (Field Programmable Gate Array), DSP (Digital Signal Processor), ASIC (Application Specific Integrated Circuit), SOC (System on Chip, SoC), etc. The above-mentioned integrated circuit or chip can be used to execute executable instructions (or codes) to implement the above-mentioned ECG signal detection method. The executable instructions can be stored in the integrated circuit or chip, or can be obtained from other devices or equipment, such as the integrated circuit or chip including a processor, a memory, and an interface for communicating with other devices. The executable instruction can be stored in the memory, and when the executable instruction is executed by the processor, the above-mentioned method for detecting the ECG signal can be implemented; alternatively, the integrated circuit or chip can receive the executable instruction through the interface and transmit it to the processor for execution, so as to implement the above-mentioned method.
[0216] In another exemplary embodiment, a computer program product is further provided. The computer program product includes a computer program that can be executed by a programmable device. The computer program has a code portion for executing the above-mentioned electrocardiogram signal detection method when executed by the programmable device.
[0217] Fig.12 FIG. 1 is a block diagram of another device 1200 for detecting an electrocardiogram signal according to an exemplary embodiment. For example, the device 1200 may be provided as a server. Fig.12 The device 1200 includes a processing component 1222, which further includes one or more processors, and a memory resource represented by a memory 1232 for storing instructions executable by the processing component 1222, such as an application. The application stored in the memory 1232 may include one or more modules, each corresponding to a set of instructions. In addition, the processing component 1222 is configured to execute instructions to perform the above-mentioned electrocardiogram signal detection method.
[0218] The device 1200 may also include a power supply component 1226 configured to perform power management of the device 1200, a wired or wireless network interface 1250 configured to connect the device 1200 to a network, and an input / output interface 1258. The device 1200 may operate based on an operating system stored in the memory 1232.
[0219] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the present disclosure. The present disclosure is intended to cover any variations, uses or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary techniques in the art that are not disclosed in the present disclosure. The description and examples are to be considered as exemplary only, and the true scope and spirit of the present disclosure are indicated by the following claims.
[0220] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.
Claims
1. A method for detecting an electrocardiogram signal, characterized in that: include: Acquire an initial electrocardiogram signal, and determine a target start and end position and a third-order derivative of the initial electrocardiogram signal according to the initial electrocardiogram signal; Positioning a target pacing pulse according to the third-order derivative within the range where the target start and end positions are located; Performing feature enhancement processing on the initial ECG signal according to the target pacing pulse to obtain a target ECG signal; The target ECG signal is input into a pre-trained residual network model, so that the pre-trained residual network model outputs a detection result of the target ECG signal.
2. The method for detecting an electrocardiogram signal according to claim 1, characterized in that: The step of determining the target start and end positions and the third-order derivative of the initial electrocardiogram signal according to the initial electrocardiogram signal comprises: Performing head-to-tail splicing processing on the initial electrocardiogram signal to obtain first sampling data, and determining the target start and end positions of the initial electrocardiogram signal according to the first sampling data; The initial electrocardiogram signal is calculated to obtain a third-order derivative of the initial electrocardiogram signal.
3. The method for detecting an electrocardiogram signal according to claim 2, characterized in that: The step of performing head-to-tail concatenation processing on the initial electrocardiogram signal to obtain first sampling data includes: Performing interception processing on the initial electrocardiogram signal to obtain second sampling data, and performing inversion processing on the second sampling data to obtain third sampling data; The third sampling data and the initial electrocardiogram signal are spliced to obtain the first sampling data.
4. The method for detecting an electrocardiogram signal according to claim 2, characterized in that: The step of determining the target start and end positions of the initial electrocardiogram signal according to the first sampling data includes: Traversing the first sampling data to determine a peak position in the first sampling data, and determining a peak interval according to the peak position; The initial start and end positions of the initial electrocardiogram signal are determined according to the peak position and the peak interval, and the initial start and end positions are filtered to obtain target start and end positions.
5. The method for detecting an electrocardiogram signal according to claim 1, characterized in that: The positioning of the target pacing pulse according to the third-order derivative within the range where the target start and end positions are located comprises: Determining a target derivative in the third-order derivative within a range where the target start and end positions are located; Determining a derivative mean of the third-order derivative according to the target derivative, and determining a target mean template according to the derivative mean; When the target mean template and the derivative mean satisfy a preset condition, the target derivative corresponding to the derivative mean is determined as a target pacing pulse.
6. The method for detecting an electrocardiogram signal according to claim 5, characterized in that: Determining a target mean template according to the derivative mean includes: Determining a first mean template in the derivative mean, and determining a second mean template based on the derivative mean; The first mean template and the second mean template are compared to obtain a comparison result, and a target mean template is determined in the first mean template and the second mean template according to the comparison result.
7. The method for detecting an electrocardiogram signal according to claim 1, characterized in that: The step of performing feature enhancement processing on the initial electrocardiogram signal according to the target pacing pulse to obtain a target electrocardiogram signal includes: performing a zeroing process on the third-order derivative according to the target pacing pulse to obtain a filtered third-order derivative; Acquire a first amplitude corresponding to the initial electrocardiogram signal and a second amplitude corresponding to the filtered third-order derivative, and calculate the first amplitude and the second amplitude to obtain an enhancement coefficient; The initial electrocardiogram signal is subjected to feature enhancement processing according to the enhancement coefficient and the filtered third-order derivative to obtain a target electrocardiogram signal.
8. A device for detecting an electrocardiogram signal, characterized in that: include: A signal acquisition module, configured to acquire an initial ECG signal, and determine a target start and end position and a third-order derivative of the initial ECG signal according to the initial ECG signal; A pulse positioning module, configured to position a target pacing pulse according to the third-order derivative within the range where the target start and end positions are located; A feature enhancement module, configured to perform feature enhancement processing on the initial ECG signal according to the target pacing pulse to obtain a target ECG signal; The model detection module is configured to input the target ECG signal into a pre-trained residual network model so that the pre-trained residual network model outputs a detection result of the target ECG signal.
9. A computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the program instructions are executed by a processor, the steps of the method described in any one of claims 1 to 7 are implemented.
10. An electronic device, characterized in that: include: a memory having a computer program stored thereon; A processor, configured to execute the computer program in the memory to implement the steps of the method according to any one of claims 1 to 7.
