Electrocardiosignal processing method and system

Abstract

The invention discloses an electrocardiosignal processing method and system, and relates to the field of biomedical signal processing.The method comprises the steps that electrocardiosignals are obtained, heart beat cutting is conducted on the electrocardiosignals, multiple heart beats are obtained, and position information of QRS waves, T waves and ST wave bands of each heart beat is obtained through a Pan-Tompkins algorithm; extracting dynamic characteristics of each heart beat based on the position information; and evaluating the importance of the dynamic features by using a random forest algorithm, screening key features, and carrying out series fusion to obtain comprehensive feature information. According to the method, through multi-dimensional feature extraction and fusion, subtle changes in the complex electrocardiosignals are captured more comprehensively, and therefore the accuracy and efficiency of electrocardiosignal analysis are improved.

Description

Technical Field

[0001] This application relates to the field of biomedical signal processing, and particularly to a method and system for processing electrocardiogram (ECG) signals. Background Art

[0002] Early diagnosis and treatment of cardiovascular diseases are of great significance for reducing mortality and improving the quality of life of patients. As an important indicator reflecting cardiac electrical activity, the analysis of ECG signals plays a key role in the diagnosis of cardiovascular diseases.

[0003] Currently, the analysis of ECG signals mainly relies on traditional morphological indexes, such as waveform morphology, vector direction, etc. Although these methods can provide certain information about cardiac electrical activity, they have limitations in dealing with complex and variable ECG signals. Related technologies mainly rely on basic features such as waveform angles and amplitudes, which are difficult to comprehensively capture the complexity and diversity of ECG signals, restricting the extraction and analysis of in-depth information about cardiac electrical activity.

[0004] Therefore, there is an urgent need for a method for processing ECG signals that can more comprehensively and accurately extract valuable information in cardiac electrical activity. Summary of the Invention

[0005] The purpose of this application is to provide a method and system for processing ECG signals, which can more comprehensively and accurately extract key features related to cardiac electrical activity, and effectively fuse these features to reflect the complexity and diversity of cardiac electrical activity.

[0006] To achieve the above purpose, this application provides the following solutions: In a first aspect, this application provides a method for processing ECG signals, including: Obtaining an ECG signal based on the Frank lead system; Performing heartbeat segmentation on the ECG signal to obtain multiple heartbeats, and obtaining the position information of the QRS wave, T wave, and ST segment of each heartbeat through the Pan-Tompkins algorithm; each heartbeat contains a preset number of sampling points; Based on the position information of the QRS wave, T wave, and ST segment of each heartbeat, respectively extracting the dynamic features of each heartbeat; the dynamic features include multi-lead characteristic wave features, planar projection features, and three-dimensional space features; Calculating the importance score of each dynamic feature relative to all the dynamic features of the ECG signal through the random forest algorithm, and taking the dynamic features corresponding to the importance score greater than the preset importance score threshold as the key dynamic features; Based on the importance scores corresponding to the key dynamic features, performing serial fusion on the key dynamic features to obtain the comprehensive feature information of the ECG signal.

[0007] Second aspect, the present application provides a processing system for electrocardiogram signals, including: A signal acquisition module, configured to acquire electrocardiogram signals based on the Frank lead system; A heartbeat segmentation and characteristic wave detection module, configured to segment electrocardiogram signals to obtain multiple heartbeats, and obtain the position information of the QRS wave, T wave, and ST segment of each heartbeat through the Pan-Tompkins algorithm; each heartbeat contains a preset number of sampling points; A feature extraction module, configured to extract the dynamic features of each heartbeat respectively based on the position information of the QRS wave, T wave, and ST segment of each heartbeat; the dynamic features include multi-lead characteristic wave features, planar projection features, and three-dimensional space features; A feature screening module, configured to calculate the importance score of each dynamic feature relative to all the dynamic features of the electrocardiogram signal through the random forest algorithm, and use the dynamic features corresponding to the importance score greater than the preset importance score threshold as the key dynamic features; A feature fusion module, configured to perform serial fusion on the key dynamic features based on the importance scores corresponding to the key dynamic features to obtain the comprehensive feature information of the electrocardiogram signal.

[0008] According to the specific embodiments provided by the present application, the present application has the following technical effects: The present application provides a method and system for processing electrocardiogram signals. By using the Pan-Tompkins algorithm to obtain the R-wave peak position information of electrocardiogram signals, the problem that it is difficult to accurately locate the R wave in the preprocessing stage of traditional methods for electrocardiogram signals is solved, and accurate heartbeat segmentation of electrocardiogram signals is realized. By improving the Pan-Tompkins algorithm and adding a preset sliding window, a preset amplitude threshold, and a preset slope threshold, the accuracy problem in obtaining the position information of the T wave and ST segment is solved, and more accurate identification of the position information of the QRS wave, T wave, and ST segment is realized. In addition, based on the position information of the QRS wave, T wave, and ST segment of each heartbeat, dynamic features including multi-lead characteristic wave features, planar projection features, and three-dimensional space features are extracted, solving the problem of incomplete feature extraction in the prior art and realizing a more comprehensive description of cardiac electrical activity. Finally, by calculating the feature importance score through the random forest algorithm and performing serial fusion on the key dynamic features based on the score, the problems of feature selection and fusion are solved, and the efficient extraction of the comprehensive feature information of electrocardiogram signals is realized, providing new technical support for the in-depth analysis of cardiac electrical activity. Description of the Drawings

[0009] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the accompanying drawings required in the embodiments. Obviously, the accompanying drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can also be obtained based on these drawings.

[0010] Figure 1 It is an application environment diagram of a method for processing electrocardiogram signals in an embodiment of the present application.

[0011] Figure 2 It is a schematic flowchart of a method for processing electrocardiogram signals provided in an embodiment of the present application.

[0012] Figure 3 It is a schematic diagram of a single heartbeat waveform based on the Frank lead system provided in an embodiment of the present application.

[0013] Figure 4 It is a three-dimensional space loop diagram and a two-dimensional projection plane diagram of an electrocardiogram vector loop provided in an embodiment of the present application; wherein, Figure 4 (a), Figure 4 (b) and Figure 4 (c) are respectively schematic diagrams of the amplitudes of the X, Y, and Z leads of the electrocardiogram vector loop; Figure 4 (d) is a three-dimensional space loop diagram and a two-dimensional projection plane diagram of the electrocardiogram vector loop.

[0014] Figure 5 It is a diagram of electrocardiogram signal processing and characteristic wave detection provided in an embodiment of the present application.

[0015] Figure 6 It is a complete heartbeat waveform detection diagram provided in an embodiment of the present application; wherein, Figure 6 (a) is a complete heartbeat waveform detection diagram of a healthy individual; Figure 6 (b) is a complete heartbeat waveform detection diagram of a myocardial infarction patient.

[0016] Figure 7 It is a schematic diagram of the functional modules of a system for processing electrocardiogram signals provided in an embodiment of the present application.

[0017] Figure 8 It is a schematic diagram of the structure of a computer device provided in an embodiment of the present application. Detailed implementation manners

[0018] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts shall fall within the protection scope of the present application.

[0019] To make the above objects, features, and advantages of the present application more obvious and understandable, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0020] The method for processing electrocardiogram signals provided in the embodiments of the present application can be applied to an application environment as Figure 1 shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be set separately, integrated on the server 104, or placed on the cloud or other servers. The terminal 102 can send the electrocardiogram signals obtained based on the Frank lead system to the server 104. After receiving the electrocardiogram signals, the server 104 obtains the electrocardiogram signals; performs heartbeat segmentation on the electrocardiogram signals to obtain multiple heartbeats, and obtains the position information of the QRS wave, T wave, and ST segment of each heartbeat through the Pan-Tompkins algorithm; each heartbeat contains a preset number of sampling points; based on the position information of the QRS wave, T wave, and ST segment of each heartbeat, the dynamic characteristics of each heartbeat are extracted respectively; the dynamic characteristics include multi-lead characteristic wave characteristics, planar projection characteristics, and three-dimensional space characteristics; calculates the importance score of each dynamic characteristic relative to all the dynamic characteristics of the electrocardiogram signal through the random forest algorithm, and takes the dynamic characteristics corresponding to the importance score greater than the preset importance score threshold as the key dynamic characteristics; based on the importance scores corresponding to the key dynamic characteristics, performs serial fusion on the key dynamic characteristics to obtain the comprehensive characteristic information of the electrocardiogram signal. The server 104 can feedback the obtained comprehensive characteristic information of the electrocardiogram signal to the terminal 102. In addition, in some embodiments, the method for processing electrocardiogram signals can also be implemented independently by the server 104 or the terminal 102. For example, the terminal 102 can directly process the electrocardiogram signals to be processed, or the server 104 can obtain the electrocardiogram signals to be processed from the data storage system and process the electrocardiogram signals to be processed.

[0021] Among them, the terminal 102 can be but is not limited to various desktop computers, laptop computers, smart phones, tablet computers, Internet of Things devices, and portable wearable devices. The Internet of Things devices can be smart speakers, smart TVs, smart air conditioners, smart in-vehicle devices, etc. The portable wearable devices can be smart watches, smart bracelets, head-mounted devices, etc. The server 104 can be implemented by an independent server or a server cluster composed of multiple servers, and can also be a cloud server.

[0022] In an exemplary embodiment, as Figure 2 shown, a method for processing electrocardiogram signals is provided. This method is executed by a computer device, and specifically can be executed alone by a computer device such as a terminal or a server, or jointly executed by a terminal and a server. In the embodiments of the present application, taking this method applied to Figure 1 the server 104 in as an example for illustration, it includes the following steps 201 to step 205. Among them: Step 201, obtain electrocardiogram signals based on the Frank lead system.

[0023] Step 202, perform heartbeat segmentation on the electrocardiogram signals to obtain multiple heartbeats, and obtain the position information of the QRS wave, T wave, and ST segment of each heartbeat through the Pan-Tompkins algorithm; each heartbeat contains a preset number of sampling points.

[0024] Step 203, based on the position information of the QRS wave, T wave, and ST segment of each heartbeat, extract the dynamic characteristics of each heartbeat respectively; the dynamic characteristics include multi-lead characteristic wave characteristics, planar projection characteristics, and three-dimensional space characteristics.

[0025] Step 204, calculate the importance score of each dynamic characteristic relative to all the dynamic characteristics of the electrocardiogram signal through the random forest algorithm, and use the dynamic characteristics corresponding to the importance score greater than the preset importance score threshold as the key dynamic characteristics.

[0026] Step 205, based on the importance scores corresponding to the key dynamic characteristics, perform serial fusion on the key dynamic characteristics to obtain the comprehensive characteristic information of the electrocardiogram signal.

[0027] By implementing the above steps 201 to 205, the present application can effectively extract and fuse key kinetic features from the electrocardiogram (ECG) signals, thereby improving the accuracy and reliability of ECG signal analysis. By using the Pan-Tompkins algorithm and its improved versions to precisely locate the key waveforms in the ECG signals, the present application can capture the complex changes in cardiac electrical activities more comprehensively. In addition, by using the random forest algorithm to evaluate and screen the importance of features, the present application ensures that only the most relevant features are used for the final feature fusion, which helps to reduce noise and improve the diagnostic accuracy. Finally, by serially fusing the key kinetic features, the present application provides a comprehensive feature information that can be used for more accurate analysis and diagnosis of cardiovascular diseases, providing strong support for clinical decision-making.

[0028] In another exemplary embodiment of the present application, before step 202, it further includes: performing denoising processing on the ECG signal by using the discrete wavelet transform method.

[0029] In another exemplary embodiment of the present application, step 202 specifically includes: Step 301, performing median filtering processing and band-pass filtering processing on the heartbeat signal of each heartbeat to obtain the denoised heartbeat signal of each heartbeat.

[0030] As Figure 5 shown, in the ECG signal processing, median filtering and discrete wavelet decomposition methods are used to perform denoising processing on the ECG signal to effectively remove noise interference. Subsequently, the Pan-Tompkins algorithm is used to locate the R-wave peaks of the denoised ECG signal, and the signal is segmented into heartbeats according to the R-wave peak positions. Specifically, with each R-wave peak as the center, 250 sampling points are taken forward and 400 sampling points are taken backward, so as to obtain a complete heartbeat containing 651 sampling points. This method can effectively retain the morphological features of the ECG signal and provide a high-quality data basis for subsequent heartbeat analysis and feature extraction.

[0031] In another exemplary embodiment of the present application, as Figure 5 shown, median and band-pass filtering are performed on the ECG signal to remove low-frequency and high-frequency noises and improve the signal quality. The band-pass filter removes low-frequency and high-frequency noises, thereby retaining the main signal components of the QRS complex. Generally, the frequency range of the band-pass filter is set to 0.5 Hz to 50 Hz, where low-frequency noises (such as baseline drift) generally appear below 0.5 Hz, while high-frequency noises (such as electromyographic noise) usually appear above 50 Hz. The signal passes through a median filter with cut-off frequencies of 5 Hz and 100 Hz to attenuate the T wave and P wave, and the shape of the QRS complex in the signal is presented more clearly.

[0032] Step 302: After performing differential operation and squaring operation on the denoised heartbeat signal of each heartbeat, a squared heartbeat signal is obtained.

[0033] Step 303: Determine the R-wave peak position as the position corresponding to the sampling points in the squared heartbeat signal that are greater than the preset R-wave peak amplitude threshold and within the preset time threshold range.

[0034] Step 304: Determine the starting point position of the Q wave as the first local minimum before the R-wave peak position; determine the starting point position of the S wave as the first local minimum after the R-wave peak; determine the ending point position of the S wave as the first local minimum after the first local maximum after the starting point position of the S wave.

[0035] Using differential operation to highlight the rapid changes of the QRS complex for more accurate detection of the QRS complex; squaring operation to remove negative values in the signal so that all wave peaks are positive, thereby enhancing the peak of the R wave; the Q wave is the first local minimum before the R wave, the S wave is the second local minimum after the R wave, and within a specified time window after the R wave, perform S wave detection to limit the search range to effectively identify the accurate position of the S wave; adaptively detect the starting point and ending point of the T wave according to the position and timing characteristics of the QRS complex. By analyzing the duration ratio relationship of the QRS complex and T wave processes, determine the ending point position of the T wave, and determine the ST segment according to the ending point of the QRS complex and the starting point of the T wave. The schematic diagram of a single heartbeat waveform in Frank lead is as Figure 3 shown.

[0036] The reference points in the electrocardiogram signal are the positions of the peaks, starts, and offsets of the waveform. The QRS complex, ST segment, and T wave all indicate the position information of cardiac electrical activities. Therefore, they can be used for diagnosing heart diseases. Using the Pan-Tompkins algorithm to add the positioning of the adaptive T wave ending on the basis of the already achieved QRS complex positioning to complete the preprocessing of the electrocardiogram signal can accurately detect the characteristic waves of the three leads in Frank lead.

[0037] Step 305: Determine the minimum duration and maximum duration of the T wave.

[0038] As an optional implementation manner, the calculation formulas for the minimum duration and maximum duration of the T wave are respectively: .

[0039] .

[0040] Where, represents the minimum duration of the T wave; represents the actual total duration of the QRS complex; represents the maximum upper limit value of the QRS complex duration; Indicates the minimum duration of the initial T wave; Indicates the maximum duration of the T wave; Indicates the maximum duration of the initial T wave.

[0041] Step 306: Use the end position of the S wave as the starting point of the first preset sliding window, and use the positions corresponding to the sampling points within the first preset sliding window that are greater than the first preset T-wave amplitude threshold and the first preset T-wave slope threshold as the starting point position of the T wave; the size of the first preset sliding window is twice the maximum duration of the T wave.

[0042] Step 307: Use the starting point position of the T wave as the midpoint of the second preset sliding window, and use the positions corresponding to the sampling points within the second preset sliding window that are greater than the second preset T-wave amplitude threshold and the second preset T-wave slope threshold as the first screening position of the T wave; the size of the second preset sliding window is twice the minimum duration of the T wave; Step 308: Use the position obtained by adding the minimum duration of the T wave to the end position of the QRS wave as the second screening position of the T wave.

[0043] Step 309: Select the position corresponding to the sampling point with the largest time among the first screening position and the second screening position as the end position of the T wave. Select the point with the later time, that is, the larger value, as the finally optimized end position of the T wave.

[0044] Step 310: Use the end position of the S wave and the starting point position of the T wave as the starting point position and the end point position of the ST segment respectively.

[0045] Based on the implementation of QRS wave localization using the Pan-Tompkins algorithm, the localization of the adaptive T wave starting point and end point is added. Subsequently, the position of the ST segment can be determined through the end point of the QRS wave and the starting point of the T wave, thereby completing the preprocessing of the electrocardiogram signal and accurately detecting the characteristic waves of the electrocardiogram signal.

[0046] Under the Pan-Tompkins algorithm framework, the QRS wave shows rapid rise and fall in cardiac electrical activity. In contrast, the changes in P waves and T waves are relatively gentle. Differentiation operations are used to enhance the rapid change characteristics of the QRS wave. Through differentiation operations, the amplitude of the QRS wave is significantly amplified, while effectively weakening the interference effects of P waves and T waves, thereby highlighting the characteristics of the QRS wave in complex electrocardiogram signals. Subsequently, a squaring operation is performed on the differentiated signal. This operation converts the negative values in the signal into positive values, making all wave peaks present positive states, significantly enhancing the peak value of the QRS wave, especially highlighting the characteristics of the R wave. The QRS wave plays a core role in the conduction process of cardiac electrical activity and has unique morphological and temporal characteristics. Based on this characteristic, the PT (Pan-Tompkins) algorithm screens the processed signal by setting time and amplitude thresholds to identify possible QRS waves.

[0047] As an alternative implementation, the T wave represents the repolarization process of the ventricles, with a normal duration of 0.10 - 0.25 seconds. There is a certain correlation between the ventricular depolarization time and the repolarization process. Therefore, in this embodiment, when it is detected that the QRS wave exceeds 0.11 seconds, the T wave adaptive adjustment is initiated.

[0048] As an alternative implementation, starting from the end point of the QRS wave, given that the duration of the ST segment is usually about 0.15 seconds, in this embodiment, 0.15 seconds after the end point of the QRS wave is selected as the starting point, and a first preset sliding window is defined by extending 0.15 seconds forward and backward respectively as the candidate range for the starting point of the T wave. Screening is performed using the first preset T wave amplitude threshold and the first preset T wave slope threshold. By statistically analyzing the signal amplitudes within the candidate range of the starting point of the T wave, calculating their mean and standard deviation, and then using the average amplitude as the first preset T wave amplitude threshold. Similarly, for the signal slope values within the statistical candidate range of the starting point of the T wave, calculating their mean and standard deviation, and setting the average slope as the first preset T wave slope threshold. By calculating the mean amplitude and amplitude standard deviation within the candidate range, the first point in the candidate region of the starting point of the T wave that comprehensively exceeds these two values is determined as the starting point of the T wave.

[0049] As an alternative implementation, the first preset T wave amplitude threshold is: .

[0050] .

[0051] Wherein, represents the first preset T wave amplitude threshold, that is, the average value of the amplitudes of all sampling points within the first preset sliding window; N represents the total number of sampling points within the first preset sliding window; represents the signal amplitude of the nth sampling point within the first preset sliding window, It represents the standard deviation of the signal amplitude and is used to measure the degree of dispersion of the electrocardiogram signal amplitude within the candidate range.

[0052] The first preset T-wave slope threshold is: .

[0053] .

[0054] Among them, It represents the first preset T-wave slope threshold, that is, the average value of the slopes of all sampling points within the first preset sliding window; It represents the slope of the nth sampling point within the first preset sliding window, It represents the standard deviation of the slope and is used to measure the degree of dispersion of the signal slope within the candidate range.

[0055] In this embodiment, after determining the T-wave starting point, in order to accurately screen out the T-wave ending point, the following optimization strategy can be adopted. Based on the minimum T-wave duration of 0.1 second, extend 0.1 second before and after the T-wave starting point, and the interval thus delimited is used as the second preset sliding window, that is, the candidate range of the T-wave ending point. Within the candidate range of the T-wave ending point, use the method based on amplitude and slope thresholds mentioned above to calculate the second preset T-wave amplitude threshold and the second preset T-wave slope threshold for the preliminary screening of the T-wave ending point. Specifically, by combining the threshold conditions set by the mean value, standard deviation of the amplitude and the relevant statistical characteristics of the slope, perform point-by-point analysis on the electrocardiogram signal within the candidate range of the T-wave ending point, so as to determine the preliminarily screened T-wave ending point.

[0056] The detection of Q, S, R, and T waves in each heartbeat is completed, and the waveform diagram of the detection result is drawn through a visualization tool, as Figure 6 shown, and the QRS wave, ST segment, and T wave are visually displayed respectively. Among them, the complete heartbeat waveform detection diagram is as Figure 6 shown; among them, Figure 6 (a) is the waveform detection diagram of a complete heartbeat signal of a healthy individual; Figure 6 (b) is the waveform detection diagram of a complete heartbeat signal of a myocardial infarction patient. Figure 6 (a) and Figure 6 (b) The leftmost yellow background part is the QRS wave, the middle orange background part is the ST segment, and the rightmost green background part is the T wave detected by the adaptive T-wave detection method. The differential and square combination method is used to locate the characteristic band, the threshold window method is combined to accurately detect the Q point and S point, and the T wave is detected through the adaptive detection method. It can be clearly seen from Figure 6 the position distribution of each wave point, verifying the effectiveness of the algorithm. By comparing Figure 6 (a) and Figure 6In (b), the differences in the heartbeat signals between healthy individuals and myocardial infarction patients can be observed, further verifying the applicability and accuracy of the method of this application in different situations.

[0057] In another exemplary embodiment of this application, the dynamic information in different dimensions of the electrocardiogram signal can reflect the dynamic changes of cardiac electrical activity. For example, the change in the maximum vector of the QRS loop exceeds the healthy range, the loop is affected by the change in cardiac electrical activity resulting in asymmetry of the electrocardiogram vector, and the disappearance of the electrocardiogram vector at a certain position will cause the complete loss of normal electrical activity in that area and then the overall electrocardiogram vector will deviate towards the place where the electrocardiogram vector disappears, thus causing changes in the dynamic information. The characteristic changes mainly focus on the QRS loop, the T loop, and the ST segment. Therefore, the lengths of the QRS loop and the T loop, the maximum vectors and angles of the QRS loop and the T loop, the included angle between the maximum vectors of the QRS loop and the T loop, the octant-related characteristics, the mean value of the ST segment, as well as the projected plane area and area ratio are extracted to display the dynamic changes of the cardiac electrocardiogram vector. And in order to evaluate the signal complexity from the QRS loop to the T loop and detect abnormal signals, the sample entropy from the QRS wave to the T wave is specifically added. The features are divided into the following three dimensions: The multi-lead characteristic wave features in step 203 specifically include: the mean value of the signal amplitudes of all sampling points in the ST segment, the peak value of the T wave, and the sample entropy from the QRS wave to the T wave.

[0058] The maximum value of the T wave, that is, the peak value of the T wave, usually represents the direction of the strongest current during the cardiac repolarization process, that is, the dominant direction of cardiac repolarization. The T wave may be inverted or flattened, and even form an inconsistent distribution on the projection plane.

[0059] 。

[0060] Among them, represents the peak value of the T wave, represents finding the maximum value of the absolute value of the T wave potential value among the th sampling points in the T wave band, which is the peak value of the T wave.

[0061] The elevation or depression of the ST segment is usually a sign of cardiac ischemia or other cardiac problems. Therefore, the mean value of the ST segment is calculated to reflect whether the heart is abnormal. The following formula is used to calculate the mean value of the signal amplitudes of all sampling points in the ST segment: 。

[0062] Among them, represents the mean value of the signal amplitudes of all sampling points in the ST segment, represents the starting point position of the ST segment, represents the ending point position of the ST segment, represents the signal amplitude of the th sampling point in the ST segment.

[0063] Sample entropy is an index to measure the complexity and self-similarity of a time series. The smaller its value, the stronger the regularity of the signal; the larger the value, the higher the complexity of the signal. The calculation of sample entropy depends on the time series of data and a tolerance value, usually denoted as r, to evaluate the number of similar patterns in the series. Sample entropy measures the complexity of a time series by comparing the matching probabilities of template vectors with lengths m and m + 1. The sample entropy from the QRS wave to the T wave is calculated using the following formula: .

[0064] Wherein, represents the sample entropy from the QRS wave to the T wave; l represents the length of the template vector; r represents the preset tolerance value; represents the total number of sampling points from the starting point of the QRS wave to the ending point of the T wave; represents under r the matching probability between any two vectors with length m from the QRS wave to the T wave; represents under r the matching probability between any two vectors with length m + 1 from the QRS wave to the T wave.

[0065] The approximate value variance is a method for estimating the variance of sample data, usually used to describe the dispersion or volatility of data. The calculation formula for the approximate variance from the QRS wave to the T wave is as follows.

[0066] .

[0067] Wherein, represents the approximate variance from the QRS wave to the T wave; represents the total number of sampling points from the starting point of the QRS wave to the ending point of the T wave, represents the th sampling point from the QRS wave to the T wave, represents the amplitude of the th sampling point from the QRS wave to the T wave, represents the average value of the amplitudes of all sampling points from the QRS wave to the T wave.

[0068] In wavelet transform, a signal is decomposed into components of different frequencies, and the detail coefficients represent high-frequency information. The variance of the detail coefficients is used to describe the variability or complexity of the high-frequency part of the signal. Suppose the sampling point sequence from the QRS wave to the T wave undergoes wavelet transform to obtain detail coefficients, then the calculation formula for the variance of the detail coefficients is as follows: .

[0069] Wherein, Represents the variance of the detail coefficients from the QRS complex to the T wave; Represents the detail coefficient of the th sampling point after the sampling point sequence from the QRS complex to the T wave undergoes wavelet transform, Represents the mean value of the detail coefficients of all sampling points after the sampling point sequence from the QRS complex to the T wave undergoes wavelet transform.

[0070] The planar projection features in step 203 specifically include: the included angle of the QT maximum vector, the maximum vector length, angle, and 12 - quadrant area ratio of the projected surface loops of the QRS loop and the T loop on the two - dimensional plane; the QRS loop is formed by the potential values of the QRS complex of each heartbeat on the X, Y, and Z leads based on the Frank lead system; the T loop is formed by the potential values of the T wave of each heartbeat on the X, Y, and Z leads based on the Frank lead system.

[0071] Formed by the potential values of the QRS complex of each heartbeat on the X, Y, and Z leads based on the Frank lead system. The specific steps are as follows: In each heartbeat, extract the potential values of the QRS complex from the X, Y, and Z leads. These potential values reflect the voltage changes of the cardiac electrical activity in the X, Y, and Z directions during the QRS complex. For each sampling point in a heartbeat, during the QRS complex, combine the potential values of the X, Y, and Z leads into a three - dimensional QRS vector; connect the three - dimensional QRS vectors of all sampling points in each heartbeat in chronological order to form a closed loop, which is the QRS loop. This loop reflects the trajectory of the electrical activity of the QRS complex in three - dimensional space during one heartbeat.

[0072] The formation process of the T loop is similar to that of the QRS loop, but focuses on the electrical activity during the T wave. In each heartbeat, extract the potential values of the T wave from the X, Y, and Z leads. These potential values reflect the voltage changes of the cardiac electrical activity in the X, Y, and Z directions during the T wave. For each sampling point in a heartbeat, during the T wave, combine the potential values of the X, Y, and Z leads into a three - dimensional T vector; then connect the three - dimensional T vectors of all sampling points in each heartbeat in chronological order to form a closed loop, which is the T loop. This loop reflects the trajectory of the electrical activity of the T wave in three - dimensional space during one heartbeat.

[0073] Compared with the QRS loop, the T loop more reflects the spatial distribution and dynamic changes of the ventricular repolarization process. By analyzing the characteristics such as the shape, size, and direction of the T loop, the synchrony of cardiac repolarization and the direction of the cardiac electrical axis can be evaluated. In addition, the morphology and direction of the T loop may be related to the physiological state and pathological changes of the heart, so it has important application value in clinical diagnosis and research.

[0074] The three - dimensional space loop diagram and its two - dimensional projection surface diagram of the electrocardiogram vector loop are asFigure 4 as shown; among which, Figure 4 (a), Figure 4 (b) and Figure 4 (c) are respectively the amplitude schematic diagrams of the X, Y, and Z leads of the electrocardiogram vector loop; Figure 4 (d) is the three-dimensional space loop diagram and two-dimensional projection plane diagram of the electrocardiogram vector loop. The electrocardiogram vector loop includes the QRS loop and the T loop. Both the QRS loop and the T loop are constructed by analyzing the potential changes of the electrocardiogram signal in three dimensions, and they respectively reflect the electrical activity trajectories during ventricular depolarization and repolarization. By comparing and analyzing the characteristics of these two loops, the electrophysiological characteristics and functional status of the heart can be more comprehensively understood, providing important reference information for the diagnosis and treatment of heart diseases.

[0075] The QT maximum vector angle is calculated using the following formula: .

[0076] Among which, represents the QT maximum vector angle; represents the cosine value of the QT maximum vector angle; V1 and V2 respectively represent the first vector composed of sampling points in the first time period between the starting point of the QRS complex and the ending point of the T wave and the second vector composed of sampling points in the second time period; and respectively represent the magnitudes of the first vector and the second vector. The QT maximum vector angle is a parameter that measures the change direction of the QT interval or the direction of electrical activity of the heart. The QT interval represents the time length of a complete electrical activity cycle of the heart (from ventricular depolarization to repolarization), and the maximum vector angle involves the change angle of the QT interval vector at different time points, describing the direction change of the heart's electrical activity from depolarization to repolarization. The change in the magnitude of the angle can, to a certain extent, reveal the electrophysiological abnormal changes in the heart during MI (Myocardial Infarction).

[0077] The Euclidean distance formula is used to calculate the maximum vector length of the projection plane loop of the QRS loop and the T loop on the two-dimensional plane.

[0078] The arctangent function is used to calculate the angle and the area ratio of the 12 quadrants of the projection plane loop.

[0079] In the projection plane of the QRS loop and the T loop on the two-dimensional plane, the maximum distance between two points is calculated, and arctan2 can be used to calculate the angles in different quadrants.

[0080] .

[0081] Among which, Indicates the maximum distance between the point and the point in the projection plane of the QRS loop and the T loop on the two-dimensional plane, that is, the Euclidean distance between the two points. And indicate the abscissa and ordinate of the point in the projection plane of the QRS loop and the T loop on the two-dimensional plane, And indicate the abscissa and ordinate of the point in the projection plane of the QRS loop and the T loop on the two-dimensional plane, indicate the point and the angle between points in the projection plane of the QRS loop and the T loop on the two-dimensional plane, and arctan2 is the arctangent function.

[0082] The 12-quadrant area ratio is a feature derived based on the doctor's clinical observation of the offset and shape change of the VCG (Vectorcardiogram) in the projection plane. The projection plane area is calculated by the Gaussian area formula, ensuring that each vertex is connected to the next vertex. In this way, the area of the irregular figure can be calculated by finding the area of triangles and then accumulating and dividing by 2.

[0083] The 12-quadrant area ratio is calculated using the following formula: .

[0084] .

[0085] Among them, represents the proportion of the area of the th quadrant in the total area of the projection plane; = 1, 2, 3....., 12; represents the area of the th quadrant; represents the total area of the projection plane; J represents the total number of sampling points in the projection plane; x0 and y0 respectively represent the abscissa and ordinate of the starting sampling point in the projection plane; x j and y j respectively represent the abscissa and ordinate of the jth sampling point in the projection plane; x j+1 and y j+1 respectively represent the abscissa and ordinate of the (j + 1)th sampling point in the projection plane.

[0086] The three-dimensional space features in step 204 specifically include: the lengths of the QRS loop and the T loop, the maximum vectors and angles; the angle between the maximum vectors of the QRS loop and the T loop; the octant values, octant time ratios, octant vector variances, and octant maximum vector lengths of the QRS loop and the T loop.

[0087] The dynamic information of the electrocardiogram (ECG) signal can reflect the dynamic changes of cardiac electrical activity, such as changes in the conduction path of the electrical signal or the length of the conduction path. These changes may manifest as adjustments in the loop morphology or changes in the spatial distribution of electrical activity. By calculating the accumulation of the lengths from each point to the next point in the loop, the length of the loop can be quantified to reflect the overall situation of cardiac electrical activity. The formula is as follows: .

[0088] Where I represents the total number of ECG vectors on the loop; , and respectively represent the coordinate components of the -th ECG vector on the X-axis, Y-axis, and Z-axis of the coordinate system; represents the maximum length of the loop; represents the -th ECG vector's modulus length.

[0089] Eight equally divided regions in space: The eight equally divided regions of the ECG vector loop are obtained by dividing the spatial projection of cardiac electrical activity into eight regions, which is convenient for analyzing the direction and intensity distribution of cardiac electrical activity. Through the analysis of the ECG vector loop in these regions, it helps to perform more detailed spatial localization and diagnosis of cardiac electrical activity. By dividing the space into eight equally divided regions, clinicians can more clearly judge the cardiac electrical activity. Extracting features such as the mean value, time ratio, vector variance, and maximum vector length in each octant region can understand the changes in cardiac electrical activity and thus judge the health of cardiac activity.

[0090] Octant mean: The mean value of each octant reflects the electrical activity intensity and distribution in a specific direction or region of the ECG vector loop. It represents the central tendency or representative value of each region, which helps to describe the data distribution in each region. When the mean value shows significant changes, the morphology of the ECG vector loop will also change, which can provide a clearer physiological state of the heart.

[0091] .

[0092] Where, represents the mean value of the -th octant, which is the average value of all ECG vector values in this region and is used to measure the central tendency of electrical activity in this region. represents the -th octant region of the ECG vector loop, ={1, 2,...., 8}, The value range of is from 1 to 8, corresponding to 8 different regions or octants. Each octant represents a specific region of cardiac electrical activity in space. represents the The total number of sampling points within a sub - body, or the number of electrocardiogram vectors contained in the th sub - body. Indicates the th electrocardiogram vector data point within the th sub - body. Specifically, ranges from 1 to , that is, the sum of all data points within each sub - body is calculated, Indicates the th electrocardiogram vector data point within the th sub - body, which is the electrocardiogram vector value of the

[0093] Octant time ratio: The ratio measures the proportion of each area in the entire time period and can reflect the distribution characteristics in terms of time.

[0094] .

[0095] Among them, Indicates the time of the th sub - body, usually the duration of the electrical activity within this sub - body, Indicates the total time, which is the sum of the times for collecting the entire electrocardiogram vector loop data, Indicates the time ratio of the th sub - body, which is the ratio of a specific time period to the total time . It reflects the proportion of the th sub - body or area in the overall time.

[0096] Octant vector variance: Variance is used to measure the fluctuation of data points relative to the mean value and reflects the stability or uncertainty of each octant.

[0097] .

[0098] Among them, Indicates the vector variance of the th sub - body, Indicates the amplitude mean of the th sub - body, which is the average amplitude of all electrocardiogram vector values within this area and is used to measure the central tendency of the electrical activity in this area.

[0099] Octant maximum vector length: The maximum vector length represents the maximum extent of the data points in space and can reflect the spatial distribution range of the data points.

[0100] .

[0101] Among them, Indicates the The maximum length of the split center electrocardiogram vector, which is used to measure the maximum intensity of the electrocardiogram vector in this area. What is calculated is the th modulus of the electrocardiogram vector of the

[0102] When dealing with high-dimensional data, feature selection is a key step to improve classification accuracy and computational efficiency. This application adopts a method of random forest based on dynamic average threshold setting for feature selection. This method has strong feature evaluation ability and can effectively identify the most informative features from a large number of features. Random forest is an ensemble learning method that constructs multiple decision trees for classification or regression tasks. Each decision tree only uses part of the features of the data during training, which helps to reduce overfitting and improve the generalization ability of the model. During the classification process, random forest calculates the importance of each feature, which is determined by analyzing the contribution of each feature to the model decision-making process. Specifically, random forest evaluates the relative importance of a feature by calculating the average split gain of the feature in all trees.

[0103] In another exemplary embodiment of this application, the calculation formula for the preset importance score threshold is: .

[0104] .

[0105] .

[0106] Among them, K represents the preset importance score threshold; represents the mean of the importance scores; k represents an adjustable parameter used to control the strictness of the threshold; represents the standard deviation of the importance scores; represents the importance score of the i-th kinetic feature, and m represents the number of kinetic features.

[0107] The different-dimensional features screened by the random forest algorithm are concatenated in a specific order to form a new high-dimensional feature vector. This concatenation order is scientifically arranged according to the contribution degree of each feature to the overall electrocardiogram signal feature expression and the internal logical relationship between different-dimensional features. For features that play a key role in reflecting the change law of the electrocardiogram vector and can significantly improve the richness of feature expression, they are concatenated first. At the same time, the complementarity between different-dimensional features is fully considered to ensure that the concatenated feature vector can retain the most effective information from each dimension to the greatest extent and strengthen the comprehensive expression ability of the multi-dimensional dynamic information of the electrocardiogram signal. The formula is as follows: .

[0108] Among them, is the comprehensive feature information, are the first key kinetic feature, the second key kinetic feature, and the th key kinetic feature respectively.

[0109] The purpose of feature extraction and fusion of multi-dimensional kinetic information based on electrocardiogram signals is to improve the deficiencies and information redundancy problems of existing methods in expressing cardiac electrical activities. By extracting features expressing the kinetic information of cardiac electrical activities from three dimensions: multi-lead characteristic waves, planar projections, and three-dimensional spatial electrocardiogram vector distributions, the dynamic changes and spatial distribution characteristics of cardiac electrical activities can be comprehensively captured. Combining feature screening based on statistical distributions and tandem fusion techniques can effectively remove redundant information and retain the most discriminative features, thereby constructing a high-dimensional feature vector and enhancing the expression ability of cardiac electrical activities. Through the organic combination of multi-dimensional features, not only the limitations of traditional methods in feature extraction are made up for, but also more comprehensive and accurate technical support is provided for the in-depth analysis and research of cardiac electrical activities.

[0110] This application also provides an application scenario that applies the above electrocardiogram signal processing method. Specifically: The electrocardiogram signal processing method provided in this embodiment can be applied in a remote medical monitoring scenario. The remote medical monitoring scenario includes a data acquisition link, a data processing link, and a data analysis link; data enters the data processing link from the data acquisition link, obtains corresponding feature information after the electrocardiogram signal is processed, and enters the data analysis link. The electrocardiogram signal processing method provided in this embodiment belongs to the feature extraction sub-link in the data processing link. Specifically in the feature extraction sub-link, this method accurately identifies and extracts the position information of QRS waves, T waves, and ST segments in the electrocardiogram signal, and calculates and fuses key kinetic features to provide high-quality feature data for subsequent data analysis and clinical diagnosis.

[0111] Based on the same inventive concept, the embodiments of this application also provide an electrocardiogram signal processing system for implementing the above-mentioned electrocardiogram signal processing method. The implementation solutions for solving problems provided by this system are similar to those recorded in the above method. Therefore, the specific limitations in one or more of the following electrocardiogram signal processing system embodiments can refer to the limitations on the electrocardiogram signal processing method in the above text and will not be elaborated here.

[0112] In an exemplary embodiment, as Figure 7 shown, an electrocardiogram signal processing system is provided, including: A signal acquisition module 401, configured to acquire an electrocardiogram signal based on the Frank lead system.

[0113] The heartbeat segmentation and characteristic wave detection module 402 is configured to segment the electrocardiogram (ECG) signal to obtain multiple heartbeats, and obtain the position information of the QRS wave, T wave, and ST segment of each heartbeat through the Pan-Tompkins algorithm; each heartbeat contains a preset number of sampling points.

[0114] The feature extraction module 403 is configured to extract the dynamic features of each heartbeat respectively based on the position information of the QRS wave, T wave, and ST segment of each heartbeat; the dynamic features include multi-lead characteristic wave features, planar projection features, and three-dimensional space features.

[0115] The feature screening module 404 is configured to calculate the importance score of each dynamic feature relative to all the dynamic features of the ECG signal through the random forest algorithm, and use the dynamic features corresponding to the importance score greater than the preset importance score threshold as the key dynamic features.

[0116] The feature fusion module 405 is configured to perform serial fusion on the key dynamic features based on the importance scores corresponding to the key dynamic features to obtain the comprehensive feature information of the ECG signal.

[0117] In an exemplary embodiment, a computer device is provided. The computer device can be a server or a terminal, and its internal structure diagram can be as Figure 8 shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O), and a communication interface. Among them, the processor, the memory, and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store ECG signal processing data. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, it implements a method for processing an ECG signal.

[0118] Those skilled in the art can understand that Figure 8The structure shown is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements. In an exemplary embodiment, a computer device is provided, including a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, the steps in the above method embodiments are implemented.

[0119] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0120] In an exemplary embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0121] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use, and processing of relevant data need to comply with relevant regulations.

[0122] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, database, or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memories. Non-volatile memories can include read-only memory (ROM), magnetic tapes, floppy disks, flash memories, optical memories, high-density embedded non-volatile memories, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memories can include random access memory (RAM) or external cache memories, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0123] The databases involved in the embodiments provided in this application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the embodiments provided in this application can be general-purpose processors, central processors, graphics processors, digital signal processors, programmable logics, data processing logics based on quantum computing, etc., without limitation.

[0124] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered to be within the scope described in this specification.

[0125] In this text, specific examples are used to elaborate on the principles and implementation manners of this application. The descriptions of the above embodiments are only used to help understand the method and its core idea of this application; at the same time, for those of ordinary skill in the art, according to the idea of this application, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation to this application.

Claims

1. A method for processing electrocardiogram signals, characterized in that The method for processing the electrocardiogram (ECG) signal includes: Obtaining the ECG signal based on the Frank lead system; Performing heartbeat segmentation on the ECG signal to obtain multiple heartbeats, and obtaining the position information of the QRS wave, T wave, and ST segment of each heartbeat through the Pan-Tompkins algorithm; each heartbeat contains a preset number of sampling points; Extracting dynamic features respectively based on the position information of the QRS wave, T wave, and ST segment of each heartbeat; the dynamic features include multi-lead characteristic wave features, planar projection features, and three-dimensional space features; Calculating the importance score of each dynamic feature relative to all the dynamic features of the ECG signal through the random forest algorithm, and taking the dynamic features corresponding to the importance score greater than the preset importance score threshold as the key dynamic features; Based on the importance scores corresponding to the key dynamic features, performing serial fusion on the key dynamic features to obtain the comprehensive feature information of the ECG signal.

2. The method for processing an electrocardiogram signal according to claim 1, characterized in that, Obtaining the position information of the QRS wave, T wave, and ST segment of each heartbeat through the Pan-Tompkins algorithm, specifically including: Performing median filtering and band-pass filtering on the heartbeat signal of each heartbeat to obtain the denoised heartbeat signal of each heartbeat; Performing differential operation and square operation on the denoised heartbeat signal of each heartbeat to obtain the squared heartbeat signal; Determining the position corresponding to the sampling point in the squared heartbeat signal that is greater than the preset R-wave peak amplitude threshold and within the preset time threshold range as the R-wave peak position; Determining the first local minimum before the R-wave peak position as the Q-wave starting point position; determining the first local minimum after the R-wave peak as the S-wave starting point position; determining the first local minimum after the first local maximum after the S-wave starting point as the S-wave end position; Determining the minimum duration of the T wave and the maximum duration of the T wave; Taking the S-wave end position as the starting point of the first preset sliding window, and taking the position corresponding to the sampling point within the first preset sliding window that is greater than the first preset T-wave amplitude threshold and the first preset T-wave slope threshold as the starting point position of the T wave; the size of the first preset sliding window is twice the maximum duration of the T wave; Taking the starting point position of the T wave as the midpoint of the second preset sliding window, and taking the position corresponding to the sampling point within the second preset sliding window that is greater than the second preset T-wave amplitude threshold and the second preset T-wave slope threshold as the first screening position of the T wave; the size of the second preset sliding window is twice the minimum duration of the T wave; Taking the position obtained by adding the minimum duration of the T wave to the QRS-wave end position as the second screening position of the T wave; Selecting the position corresponding to the sampling point with the largest time among the first screening position and the second screening position as the end position of the T wave; Taking the S-wave end position and the T-wave starting point position as the starting point position and the end point position of the ST segment respectively.

3. The method for processing an electrocardiogram signal according to claim 2, wherein The calculation formulas for the minimum duration of the T wave and the maximum duration of the T wave are respectively: ； ； Among them, represents the minimum duration of the T wave; represents the actual total duration of the QRS complex; represents the maximum upper limit value of the QRS complex duration; represents the initial minimum duration of the T wave; represents the maximum duration of the T wave; represents the initial maximum duration of the T wave.

4. The method for processing an electrocardiogram signal according to claim 2, wherein, The first preset T-wave amplitude threshold is: ； Wherein, represents the first preset T-wave amplitude threshold; N represents the total number of sampling points within the first preset sliding window; represents the signal amplitude of the nth sampling point within the first preset sliding window; The first preset T-wave slope threshold is: ； Among them, represents the first preset T-wave slope threshold; represents the slope of the nth sampling point within the first preset sliding window.

5. The processing method of the electrocardiogram signal according to claim 1, wherein, The multi-lead characteristic wave features specifically include: the mean signal amplitude of all sampling points of the ST segment, the T-wave peak, and the sample entropy from the QRS wave to the T wave.

6. The method for processing an electrocardiogram signal according to claim 1, wherein, The planar projection features specifically include: the maximum vector angle of QT, the maximum vector length, angle, and 12-quadrant area ratio of the projection surface loops of the QRS loop and the T loop on the two-dimensional plane; the QRS loop is formed by the potential values of the QRS waves of each heartbeat on the X, Y, and Z leads based on the Frank lead system; the T loop is formed by the potential values of the T waves of each heartbeat on the X, Y, and Z leads based on the Frank lead system.

7. The method for processing an electrocardiogram signal according to claim 1, wherein The three-dimensional space features specifically include: the length, maximum vector, and angle of the QRS loop and the T loop; the angle between the maximum vectors of the QRS loop and the T loop; the octant values, octant time ratios, octant vector variances, and octant maximum vector lengths of the QRS loop and the T loop.

8. The processing method of the electrocardiogram signal according to claim 1, characterized in that The calculation formula for the preset importance score threshold is: ； ； ； Among them, K represents a preset importance score threshold; represents the mean importance score; k represents an adjustable parameter; represents the standard deviation of the importance score; represents the importance score of the i-th kinetic feature, and m represents the number of kinetic features.

9. The processing method of the electrocardiogram signal according to claim 1, characterized in that, Before performing heartbeat segmentation on the electrocardiogram (ECG) signal to obtain multiple heartbeats, it also includes: Using the discrete wavelet transform method to denoise the ECG signal.

10. A processing system for electrocardiogram signals, characterized in that, The ECG signal processing system applies the ECG signal processing method described in any one of claims 1-9. The ECG signal processing system includes: A signal acquisition module for acquiring the ECG signal based on the Frank lead system; A heartbeat segmentation and characteristic wave detection module for performing heartbeat segmentation on the ECG signal to obtain multiple heartbeats, and obtaining the position information of the QRS wave, T wave, and ST segment of each heartbeat through the Pan-Tompkins algorithm; each heartbeat contains a preset number of sampling points; A feature extraction module for respectively extracting the dynamic features of each heartbeat based on the position information of the QRS wave, T wave, and ST segment of each heartbeat; the dynamic features include multi-lead characteristic wave features, planar projection features, and three-dimensional space features; A feature screening module for calculating the importance score of each dynamic feature relative to all the dynamic features of the ECG signal through the random forest algorithm, and taking the dynamic features corresponding to those greater than the preset importance score threshold as key dynamic features; A feature fusion module for performing serial fusion on the key dynamic features based on the importance scores corresponding to the key dynamic features to obtain the comprehensive feature information of the ECG signal.

Images