ECG perception signal encoding method and system

Abstract

The present invention discloses an ECG perception signal encoding method and system, which includes: segmenting multiple unlabeled ECG signals to obtain multiple sample signals; selecting a normal sample signal as a reference signal, performing a difference calculation between each sample signal and the reference signal, comparing the difference calculation result with a preset threshold, and obtaining a perception representation of each sample signal; maintaining the original sample signal corresponding to the normal perception representation according to the type of perception representation, and feeding back the abnormal perception representation to the corresponding original sample signal to obtain an ECG perception signal for each sample signal; inputting the abnormal perception representation into an autoencoder network, and outputting the encoded ECG perception signal. In an unsupervised environment, the present invention uses a normal ECG signal as a reference signal to amplify the difference between the signal with abnormal perception and the reference signal, which is beneficial to feature capture of subsequent intelligent algorithms and lays the foundation for accurate identification of electrocardiogram signal categories.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent data learning, and in particular relates to an ECG perception signal encoding method and system. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] In recent years, with the vigorous development of smart healthcare, a large number of ECG automatic analysis and diagnosis software or platforms based on artificial intelligence, big data, and Internet of Things technologies have emerged. They can support data collection, remote diagnosis, and hierarchical diagnosis and treatment, and share ECG data and reports, which can assist in the construction of hospital ECG network informatization. However, the ECG signal (i.e., ECG signal) displayed by the electrocardiogram is usually accompanied by noise such as electrode motion artifacts, muscle artifacts, and baseline drift, which affects the shape characteristics of the ECG signal. For example, the amplitude and baseline increase the difficulty of depicting the ECG signal. Therefore, it is necessary to perform noise reduction on the ECG signal before performing intelligent analysis and recognition. However, existing signal denoising methods often use wavelet or filter transforms to give high-frequency signals higher weights and low-frequency signals lower weights, resulting in the low-frequency band characteristics of the signal being difficult to be captured by intelligent algorithms, which directly affects the subsequent detection accuracy.

[0004] A normal electrocardiogram is a set of waveform curves with positive and negative waves, which can be recorded on special recording paper or displayed on an electrocardiogram oscilloscope. Figure 1 The figure shows a normal ECG signal within a cycle, which includes the P wave, QRS complex, and T wave. The P wave, QRS complex, and T wave have a certain regularity and occur in sequence. If the heart beats steadily, a typical ECG signal (i.e., a normal ECG signal) will be generated: First, the first peak (P wave) shows how the electrical impulse (excitation) propagates through the heart's two atria: the atria contract (compress), pumping blood into the ventricles, and then immediately relax; then, the electrical impulse reaches the ventricles, which can be observed in the Q, R, and S waves of the ECG, which are called the QRS complex; finally, the ventricles contract, and the T wave shows that the electrical impulse has stopped propagating, and the ventricles relax again. Through the ECG, diseases such as heart disease and arrhythmia can be detected or observed.

[0005] From the normal ECG signal displayed by the ECG, it can be seen that the QRS wave feature in the ECG signal is the most obvious, which is conducive to the feature extraction of the intelligent algorithm. However, when the ST segment in the ECG signal is abnormal, such as Figure 2As shown, intelligent algorithms have difficulty capturing these perceptual features because they vary very little across the entire ECG signal and are not very distinct. These features are likely to be filtered out as noise using the aforementioned noise reduction methods, which further increases the difficulty of perceiving low-frequency features. Therefore, sensitive knowledge of the characteristics of each band in the ECG signal is necessary to facilitate feature capture by subsequent intelligent algorithms, significantly improving the performance of intelligent analysis and recognition algorithms.

[0006] Furthermore, existing machine learning-based intelligent analysis and recognition methods require the use of large amounts of labeled ECG signals to directly extract the features of each band of the labeled ECG signals and train them to identify normal or abnormal ECG signals. On the one hand, existing methods struggle to directly and sensitively perceive or capture the features of each band in the ECG signal. On the other hand, medical data labeling generally requires manual labeling by experts, which requires significant manpower, financial, and material resources. This is unrealistic for massive amounts of signal data. Furthermore, abnormal signals appear in various forms, making it impossible to fully capture them during data collection. This also limits the scalability of supervised learning methods. Automatic labeling algorithms are risky for medical signals; once labeling is incorrect, the consequences can be fatal. Summary of the Invention

[0007] To address the deficiencies of the above-mentioned prior art, the present invention provides an ECG perception signal encoding method and system. In an unsupervised situation, the normal ECG signal is used as a reference signal, the difference between the ECG signal with abnormal perception and the normal reference signal is amplified, and self-encoding is performed. The encoded ECG perception signal is conducive to the feature capture of subsequent intelligent algorithms, facilitates the rapid perception of deep machine learning intelligent algorithms, and lays the foundation for the accurate identification and analysis of electrocardiogram signal categories.

[0008] In a first aspect, the present disclosure provides an ECG perception signal encoding method.

[0009] An ECG perception signal encoding method, comprising:

[0010] Acquire a signal set including multiple unlabeled ECG signals, segment each ECG signal in the signal set, and obtain multiple ECG sample signals;

[0011] Selecting a normal ECG sample signal as a reference signal, performing a difference calculation between each ECG sample signal and the reference signal, and comparing the difference calculation result with a preset threshold to obtain a perceptual representation of each ECG sample signal;

[0012] According to the different types of perceptual representation, the original ECG sample signal corresponding to the normal perceptual representation is maintained, and the abnormal perceptual representation is fed back to the corresponding original ECG sample signal, thereby obtaining an ECG perceptual signal for each ECG sample signal;

[0013] The ECG perception signal of each ECG sample signal is input into the autoencoder network, and the encoded ECG perception signal is output.

[0014] A further technical solution is that the segmentation is:

[0015] For each ECG signal, the R peak point in the ECG signal is selected as the starting point of the ECG sample signal, and the same time interval is intercepted backward to obtain multiple ECG sample signals to form a preprocessed ECG sample signal set.

[0016] A further technical solution, wherein the difference calculation result is compared with a preset threshold to obtain a perceptual representation of each ECG sample signal, includes:

[0017] If the sum of the absolute values ​​of the difference calculation results between the ECG sample signal and the reference signal is less than a preset threshold, the perceptual representation of the ECG sample signal is the original ECG sample signal; otherwise, the perceptual representation of the ECG sample signal is the difference calculation result between the ECG sample signal and the reference signal.

[0018] According to a further technical solution, the types of the perceptual representation include abnormal perceptual representation and normal perceptual representation, the normal perceptual representation is the original ECG sample signal, and the abnormal perceptual representation is the difference calculation result between the ECG sample signal and the reference signal.

[0019] A further technical solution, wherein the abnormal perception representation is fed back to the corresponding original ECG sample signal, includes:

[0020] Amplify and weight the abnormal perception representation, wherein the weight of the abnormal perception representation is calculated using the information gain strategy;

[0021] The weighted abnormality perception representation is fed back to the corresponding original ECG sample signal to obtain an ECG perception signal having the ECG sample signal with the perception abnormality.

[0022] In a second aspect, the present disclosure provides an ECG perception signal encoding system.

[0023] An ECG sensing signal encoding system, comprising:

[0024] An ECG signal acquisition and processing module is used to acquire a signal set containing multiple unlabeled ECG signals, segment each ECG signal in the signal set, and obtain multiple ECG sample signals;

[0025] a perceptual characterization module, configured to select a normal ECG sample signal as a reference signal, perform a difference calculation between each ECG sample signal and the reference signal, and compare the difference calculation result with a preset threshold value to obtain a perceptual characterization of each ECG sample signal;

[0026] An ECG perception signal acquisition module is used to maintain the original ECG sample signal corresponding to the normal perception representation according to the different perception representation types, and feed back the abnormal perception representation to the corresponding original ECG sample signal, thereby obtaining an ECG perception signal for each ECG sample signal;

[0027] The ECG perception signal encoding module is used to input the ECG perception signal of each ECG sample signal into the autoencoder network and output the encoded ECG perception signal.

[0028] In a third aspect, the present disclosure further provides an electronic device comprising a memory and a processor, and computer instructions stored in the memory and executed on the processor, wherein when the computer instructions are executed by the processor, the steps of the method described in the first aspect are completed.

[0029] In a fourth aspect, the present disclosure further provides a computer-readable storage medium for storing computer instructions, which, when executed by a processor, complete the steps of the method described in the first aspect.

[0030] One or more of the above technical solutions have the following beneficial effects:

[0031] 1. The present invention provides an ECG perception signal encoding method and system. In an unsupervised environment, the normal ECG signal is used as a reference signal, the difference between the ECG signal with perception abnormality and the normal reference signal is amplified, and self-encoding is performed. The encoded ECG perception signal is conducive to the feature capture of subsequent intelligent algorithms, facilitates the rapid perception of deep machine learning intelligent algorithms, and lays the foundation for the accurate identification and analysis of electrocardiogram signal categories.

[0032] 2. In the present invention, for ECG signals with lesion targets, the key targets of the lesions are highlighted by comparing them with normal ECG signals, while the normal ECG signals are not processed. In this way, the perception ability of the subsequent network classifier can be effectively enhanced, and the recognition rate of different types of ECG signals can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] The accompanying drawings, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0034] Figure 1This is a schematic diagram of the structure of a normal ECG signal within one cycle;

[0035] Figure 2 This is a schematic diagram of normal and abnormal ST segments in an electrocardiogram;

[0036] Figure 3 A flowchart for generating a perceptual representation of an abnormal ECG signal according to an embodiment of the present invention;

[0037] Figure 4 This is a comparison chart of the 1-DCNN retrieval results of the perceptually coded signal and the original signal on the MIT-BIH dataset in an embodiment of the present invention. DETAILED DESCRIPTION

[0038] It should be noted that the following detailed descriptions are exemplary and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present invention belongs.

[0039] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof.

[0040] Example 1

[0041] In order to solve the problem that existing intelligent ECG signal recognition methods require a large amount of manual labeling and the direct capture or extraction of signal features is not obvious, resulting in poor final recognition effect, this embodiment provides an ECG perception signal encoding method. Before using the intelligent recognition algorithm for recognition, the ECG perception signal is encoded to facilitate the feature capture of the subsequent intelligent algorithm, wherein the ECG perception signal refers to the signal segment in the ECG signal (i.e., the electrocardiogram signal) that plays a key role in classification or has key features for classification, and the perception is machine-oriented or algorithm-oriented. The ECG perception signal encoding method described in this embodiment specifically includes the following steps:

[0042] Step S1, obtaining a signal set including multiple unlabeled ECG signals, segmenting each ECG signal in the signal set to obtain multiple ECG sample signals;

[0043] Step S2: selecting a normal ECG sample signal as a reference signal, performing a difference calculation between each ECG sample signal and the reference signal, and comparing the difference calculation result with a preset threshold to obtain a perceptual representation of each ECG sample signal;

[0044] Step S3: maintaining the original ECG sample signal corresponding to the normal perception representation according to the different perception representation types, and feeding back the abnormal perception representation to the corresponding original ECG sample signal, thereby obtaining an ECG perception signal for each ECG sample signal;

[0045] Step S4: input the ECG perception signal of each ECG sample signal into the autoencoder network, and output the encoded ECG perception signal.

[0046] The following content further introduces the ECG perception signal encoding method proposed in this embodiment.

[0047] like Figure 3 As shown, in step S1, a signal set containing n unlabeled ECG signals is obtained. in, Represents an ECG signal of dimension m.

[0048] In this embodiment, a normal ECG signal is used as a reference signal to locate the difference between the abnormal ECG signal and the normal ECG signal, and to highlight the abnormal features in the ECG signal to facilitate the machine's rapid perception and capture. To this end, in this embodiment, each ECG signal in the signal set is first segmented, and each ECG signal is cut into from a unified peak point (R peak) to obtain the corresponding ECG sample signal. The ECG sample signal is then differenced with the normal ECG signal, amplified by an amplifier, and then sent to the autoencoder for encoding to generate a signal representation that is easy for the intelligent algorithm to quickly perceive, namely the ECG perception signal.

[0049] Specifically, for each ECG signal x i , select its R peak point as the first value of the ECG sample signal, and intercept the same time interval backward to obtain multiple ECG sample signals, forming a preprocessed ECG sample signal set. Moreover, this segmentation method makes the dimension of each ECG sample signal consistent. Assuming its dimension is d, multiple ECG sample signals are obtained by segmentation. The above time interval is the time interval of 10 seconds from the current R peak point in the ECG signal. Through this segmentation method, all extracted heart beats (ie, sample signals) have the same dimension.

[0050] In step S2, a normal ECG signal is selected from multiple ECG sample signals. As a reference signal, the difference between each sample signal and the reference signal is calculated. Assume that the perceptual representation of each sample signal is obtained as It can be characterized as:

[0051]

[0052] However, since the dataset also contains a large number of normal ECG signals, this difference may cause the perceptual representation to be highly sparse, or even close to a zero vector, thus losing its essential meaning. To this end, this embodiment introduces a preset threshold γ for control, and compares the difference calculation result with the preset threshold. If the perceptual representation of the sample signal is The sum of the absolute values ​​of If the l1 norm of the sample signal is less than the preset threshold γ, the perceptual representation of the sample signal is the original sample signal, which is taken as the normal perceptual representation. Otherwise, the operation of formula (1) is performed, that is, the perceptual representation of the sample signal is the difference calculation result between the sample signal and the reference signal, which is taken as the abnormal perceptual representation. In other words, the types of perceptual representation include abnormal perceptual representation and normal perceptual representation, where the normal perceptual representation is the original ECG sample signal and the abnormal perceptual representation is the difference calculation result between the ECG sample signal and the reference signal.

[0053] In fact, the above process is a feature re-characterization process, which is specifically characterized as follows:

[0054]

[0055] Formula (2) is used to guide the feature re-characterization of each sample signal, amplifying the signal's perception point as much as possible, while not performing any processing on the normal ECG signal. The above feature re-characterization method has the following three advantages:

[0056] (1) For normal ECG signals, since no difference processing is performed, it will be completely different from the processed signal, making it have significant heterogeneous features when it is sent to the classifier, which can enhance the recognition rate of normal ECG signals;

[0057] (2) For ECG signals with lesion targets (i.e., abnormal ECG signals), by performing a difference operation with normal signals, the key targets of the lesions can be highlighted, which can effectively enhance the perception ability of subsequent network classifiers. Although this operation changes the original physical characteristics of the ECG signal, this form of perceptual representation is more conducive to machine perception;

[0058] (3) Since the ECG signal with the lesion target only produces abnormalities in a small period of time, the feature representation obtained through difference processing is highly sparse, which can effectively reduce the impact of noise and speed up the training time of the subsequent classification algorithm.

[0059] Of course, the above method also has obvious defects, that is, after the difference operation, the ECG signal will lose its original physical meaning, and will not be readable or interpretable. For this reason, this embodiment executes step S3, that is, according to the different types of perception representation, the original ECG sample signal corresponding to the normal perception representation is maintained, and the abnormal perception representation is fed back to the corresponding original ECG sample signal to ensure the readability of the ECG signal, and the ECG perception signal of each ECG sample signal is obtained in the above manner.

[0060] The above-mentioned feedback of abnormal perception representation to the corresponding original ECG sample signal specifically includes:

[0061] First, the extracted abnormal perception representation is amplified and weighted, wherein the information gain strategy is adopted to calculate the weight of the abnormal perception representation. The information gain (ie weight) is:

[0062]

[0063] Among them, P(C j ) means C j The probability of a class signal appearing in the dataset, This means that the signal set contains abnormal perception representation The probability of This indicates abnormal perceptual representation Belongs to C j The probability of the class, Representing non-anomalous perceptual representations Belongs to C j The probability of the class, c is the number of categories.

[0064] When calculating the abnormal perception representation of ECG sample signals After the weights are determined, the weighted abnormal perception representation is fed back to the corresponding original ECG sample signal to obtain an ECG perception signal of the ECG sample signal with perception abnormality.

[0065] The above steps can be expressed by the following formula (4):

[0066]

[0067] In the above formula, if The original ECG sample signal is maintained and regarded as a normal signal by default, and only the ECG sample signal with perceived abnormality is weighted and fed back to the original ECG sample signal.

[0068] Finally, step S4 is executed to convert the ECG sensing signal of each ECG sample signal into Input into the autoencoder network, the encoding purpose of the ECG perception signal can be achieved, and the encoded ECG perception signal can be obtained. The encoded ECG perception signal obtained by the above scheme is then identified by the intelligent recognition algorithm, which can significantly improve the recognition performance. Among them, the autoencoder network is an unsupervised neural network model, which adopts the structure of encoder (Encoder)-decoder (Decoder), can learn the implicit features of the input signal, i.e. coding (Coding), and reconstruct the original signal with the learned new features, i.e. decoding (Decoding). In fact, according to different needs and application backgrounds, the autoencoder network (i.e. automatic encoder) can be divided into incomplete autoencoders, regular autoencoders, sparse autoencoders, denoising autoencoders, contraction autoencoders, etc. Since the perception signal has been amplified in the previous step, the weak perception signal will not be eliminated. Therefore, this embodiment adopts a denoising autoencoder to encode the ECG perception signal, solving the problem of filtering weak perception features by denoising the ECG signal in the prior art.

[0069] In order to further verify the effectiveness of the perceptual signal coding proposed in this embodiment, the internationally public dataset MIT-BIH is used for verification.

[0070] The MIT-BIH database consists of 48 two-lead ECG records, which include four categories: normal signal (N), left bundle branch block (L), right bundle branch block (R), and ventricular premature beats (V). This embodiment randomly selected 20,000 samples for the experiment, and randomly selected 10,000 samples as the training set and 10,000 samples as the test set. The classification algorithm selected a simple 1-DCNN for the experiment. During the experiment, the recognition results of the original signal and the data signal after perceptual encoding were compared to observe the effect of encoding. The experimental results are as follows: Figure 4 shown.

[0071] according to Figure 4 The experimental results shown in the figure show that after the data signal is perceptually encoded, the recognition rates of various categories are significantly improved, which also verifies the effectiveness of the ECG perceptual signal encoding proposed in this embodiment.

[0072] Example 2

[0073] This embodiment provides an ECG perception signal encoding system, including:

[0074] An ECG signal acquisition and processing module is used to acquire a signal set containing multiple unlabeled ECG signals, segment each ECG signal in the signal set, and obtain multiple ECG sample signals;

[0075] a perceptual characterization module, configured to select a normal ECG sample signal as a reference signal, perform a difference calculation between each ECG sample signal and the reference signal, and compare the difference calculation result with a preset threshold value to obtain a perceptual characterization of each ECG sample signal;

[0076] An ECG perception signal acquisition module is used to maintain the original ECG sample signal corresponding to the normal perception representation according to the different perception representation types, and feed back the abnormal perception representation to the corresponding original ECG sample signal, thereby obtaining an ECG perception signal for each ECG sample signal;

[0077] The ECG perception signal encoding module is used to input the ECG perception signal of each ECG sample signal into the autoencoder network and output the encoded ECG perception signal.

[0078] Example 3

[0079] This embodiment provides an electronic device, including a memory and a processor, and computer instructions stored in the memory and executed on the processor. When the computer instructions are executed by the processor, the steps in the ECG perception signal encoding method described above are completed.

[0080] Example 4

[0081] This embodiment further provides a computer-readable storage medium for storing computer instructions. When the computer instructions are executed by a processor, the steps in the ECG perception signal encoding method described above are completed.

[0082] The steps involved in the above embodiments 2 to 4 correspond to those in the method embodiment 1. For detailed implementation, please refer to the relevant description of embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media that includes one or more instruction sets; it should also be understood to include any medium that can store, encode, or carry an instruction set for execution by a processor and cause the processor to perform any method of the present invention.

[0083] Those skilled in the art will appreciate that the modules or steps of the present invention described above can be implemented using a general-purpose computer device. Alternatively, they can be implemented using program code executable by a computing device, which can then be stored in a storage device and executed by the computing device. Alternatively, they can be fabricated into separate integrated circuit modules, or multiple modules or steps can be fabricated into a single integrated circuit module for implementation. The present invention is not limited to any specific combination of hardware and software.

[0084] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be within the scope of protection of the present invention.

[0085] Although the above describes the specific embodiments of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solution of the present invention without any creative work are still within the scope of protection of the present invention.

Claims

1. A method for encoding ECG perception signals, characterized in that: include: Acquire a signal set including multiple unlabeled ECG signals, segment each ECG signal in the signal set, and obtain multiple ECG sample signals; Selecting a normal ECG sample signal as a reference signal, performing a difference calculation between each ECG sample signal and the reference signal, and comparing the difference calculation result with a preset threshold to obtain a perceptual representation of each ECG sample signal; According to the different types of perception representation, the original ECG sample signal corresponding to the normal perception representation is maintained, and the abnormal perception representation is fed back to the corresponding original ECG sample signal, thereby obtaining the ECG perception signal of each ECG sample signal, including: Amplify and weight the abnormal perception representation, wherein the weight of the abnormal perception representation is calculated using the information gain strategy; Feeding back the weighted abnormality perception representation to the corresponding original ECG sample signal to obtain an ECG perception signal having the ECG sample signal with the perceived abnormality; The ECG perception signal of each ECG sample signal is input into the autoencoder network, and the encoded ECG perception signal is output.

2. The ECG perception signal encoding method according to claim 1, wherein: The segmentation is: For each ECG signal, the R peak point in the ECG signal is selected as the starting point of the ECG sample signal, and the same time interval is intercepted backward to obtain multiple ECG sample signals to form a preprocessed ECG sample signal set.

3. The ECG perception signal encoding method according to claim 1, wherein: The step of comparing the difference calculation result with a preset threshold to obtain a perceptual representation of each ECG sample signal includes: If the sum of the absolute values ​​of the difference calculation results between the ECG sample signal and the reference signal is less than a preset threshold, the perceptual representation of the ECG sample signal is the original ECG sample signal; otherwise, the perceptual representation of the ECG sample signal is the difference calculation result between the ECG sample signal and the reference signal.

4. The ECG perception signal encoding method according to claim 1, wherein: The types of the perception representation include abnormal perception representation and normal perception representation. The normal perception representation is the original ECG sample signal, and the abnormal perception representation is the difference calculation result between the ECG sample signal and the reference signal.

5. An ECG sensing signal coding system, characterized in that: include: An ECG signal acquisition and processing module is used to acquire a signal set containing multiple unlabeled ECG signals, segment each ECG signal in the signal set, and obtain multiple ECG sample signals; a perceptual characterization module, configured to select a normal ECG sample signal as a reference signal, perform a difference calculation between each ECG sample signal and the reference signal, and compare the difference calculation result with a preset threshold value to obtain a perceptual characterization of each ECG sample signal; The ECG perception signal acquisition module is used to maintain the original ECG sample signal corresponding to the normal perception representation according to the different perception representation types, and feed back the abnormal perception representation to the corresponding original ECG sample signal, thereby obtaining the ECG perception signal of each ECG sample signal, including: Amplify and weight the abnormal perception representation, wherein the weight of the abnormal perception representation is calculated using the information gain strategy; Feeding back the weighted abnormality perception representation to the corresponding original ECG sample signal to obtain an ECG perception signal having the ECG sample signal with the perceived abnormality; The ECG perception signal encoding module is used to input the ECG perception signal of each ECG sample signal into the autoencoder network and output the encoded ECG perception signal.

6. The ECG sensing signal encoding system according to claim 5, wherein: The step of comparing the difference calculation result with a preset threshold to obtain a perceptual representation of each ECG sample signal includes: If the sum of the absolute values ​​of the difference calculation results between the ECG sample signal and the reference signal is less than a preset threshold, the perceptual representation of the ECG sample signal is the original ECG sample signal; otherwise, the perceptual representation of the ECG sample signal is the difference calculation result between the ECG sample signal and the reference signal.

7. An electronic device, characterized in that: The method comprises a memory and a processor, and computer instructions stored in the memory and executed on the processor. When the computer instructions are executed by the processor, the steps of the ECG perception signal encoding method according to any one of claims 1 to 4 are completed.

8. A computer-readable storage medium, characterized in that: Used to store computer instructions, which, when executed by a processor, complete the steps of an ECG perception signal encoding method according to any one of claims 1 to 4.

Images