An electrocardiogram monitoring method and system based on cardiac rehabilitation data

Abstract

The present invention relates to the field of electrocardiogram (ECG) monitoring technology, and in particular to an ECG monitoring method and system based on cardiac rehabilitation data. The method comprises the following steps: using a multi-lead electrocardiograph and a motion monitoring device, collecting ECG signals under different motion states, and performing fatigue ECG mapping to obtain fatigue signals. Next, analyzing the heart rate changes of the fatigue ECG mapping signal, quantifying the disordered scattered point trend of the heart rate, and fitting the dynamic inflection point distribution to obtain trend dynamic data. Based on this data, a multi-scale regular intersection analysis is performed to extract the fatigue heart rate change law, and further perform a quantitative assessment of rehabilitation deviation. Finally, the rehabilitation deviation assessment result is sent to a terminal for real-time ECG monitoring to support cardiac rehabilitation management. The present invention makes the ECG monitoring technology more perfect by optimizing the ECG monitoring technology.

Description

Technical Field

[0001] The present invention relates to the technical field of electrocardiogram monitoring, and in particular to an electrocardiogram monitoring method and system based on cardiac rehabilitation data. Background Art

[0002] Currently, ECG monitoring technology based on multi-lead ECG and motion monitoring equipment is gradually being used. It can evaluate the patient's exercise ability and rehabilitation status by comprehensively analyzing the ECG changes under different exercise states. This monitoring method based on the ECG of the exercise state can obtain the patient's ECG data under different exercise intensities in real time, and combine data analysis to dynamically evaluate the patient's cardiac load and recovery status, thereby providing a scientific basis for the formulation of personalized rehabilitation plans. However, the traditional ECG monitoring method based on cardiac rehabilitation data has the problem of large errors in trend analysis and inaccurate rehabilitation deviation analysis. Summary of the invention

[0003] Based on this, it is necessary to provide an ECG monitoring method based on cardiac rehabilitation data to solve at least one of the above technical problems.

[0004] To achieve the above object, an electrocardiogram monitoring method based on cardiac rehabilitation data comprises the following steps:

[0005] Step S1: collecting ECG signals of the patient in different motion states by means of a multi-lead electrocardiograph and a motion monitoring device to obtain ECG signals in different motion states; performing motion state fatigue ECG signal mapping on the ECG signals in different motion states to obtain fatigue ECG mapping signals;

[0006] Step S2: analyzing the heart rate variation of the fatigue ECG mapping signal to obtain fatigue heart rate variation data; performing disordered scattered point trend quantification on the fatigue heart rate variation data to obtain disordered scattered point trend quantification data of the heart rate; performing inflection point dynamic distribution fitting on the disordered scattered point trend quantification data of the heart rate to obtain trend dynamic inflection point distribution fitting data;

[0007] Step S3: performing multi-scale rule intersection analysis on fatigue heart rate change data according to trend dynamic inflection point distribution fitting data to obtain fatigue heart rate change rule intersection data; performing quantitative assessment of rehabilitation deviation based on fatigue heart rate change rule intersection data to obtain rehabilitation deviation assessment data;

[0008] Step S4: Send the rehabilitation deviation assessment data to the terminal to perform ECG monitoring of cardiac rehabilitation data.

[0009] The present invention collects the ECG signals of patients in different exercise states through a multi-lead electrocardiograph and a motion monitoring device, and can comprehensively obtain the ECG data during exercise. This process can not only reflect the impact of different exercise intensities on the heart, but also extract the signal characteristics related to fatigue by mapping the fatigue degree of the motion state of the ECG signal, and further provide accurate basic data for subsequent analysis and evaluation. This lays a solid foundation for accurately judging the patient's exercise load and fatigue degree, and is helpful for the formulation of personalized rehabilitation programs. Heart rate change analysis of fatigue ECG mapping signals can reveal the patient's heart rate fluctuations during exercise, and then reflect the heart's response and adaptability to exercise load. Through the quantification of disordered scatter trend, the disordered characteristics in the heart rate change can be extracted, thereby quantifying the stress response and recovery state of the heart. Then, the dynamic distribution fitting technology of inflection points can be applied to further reveal the dynamic inflection points and turning points that appear in the heart rate change process, which helps to deeply understand the abnormal changes in the exercise process, and then achieve accurate heart health monitoring. According to the trend dynamic inflection point distribution fitting data, multi-scale rule intersection analysis can be performed to identify the inherent rules of fatigue heart rate changes and reveal the complexity and diversity of heart rate changes under different exercise loads. Through this analysis, the key rules of fatigue heart rate changes can be extracted to provide a scientific basis for further rehabilitation evaluation. Based on the intersection data of fatigue heart rate changes, quantitative evaluation of rehabilitation deviation can not only quantify the deviation in the rehabilitation process, but also provide accurate reference for the formulation of personalized rehabilitation plans, ensuring that patients receive more scientific and accurate rehabilitation guidance. By sending the rehabilitation deviation evaluation data to the terminal, the patient's cardiac rehabilitation status can be monitored in real time. The implementation of this process ensures dynamic monitoring and feedback of heart health during the rehabilitation process, and provides real-time data support for the patient's exercise intensity and recovery. In addition, the feedback function of the terminal can help doctors and patients instantly understand the progress and deviation of cardiac rehabilitation, and then make timely adjustments during the rehabilitation process, optimize cardiac rehabilitation strategies, improve rehabilitation effects, and reduce the risk of recurrence of heart disease. Therefore, the present invention optimizes a traditional ECG monitoring method based on cardiac rehabilitation data, solves the problems of large trend analysis error and inaccurate rehabilitation deviation analysis in the traditional ECG monitoring method based on cardiac rehabilitation data, reduces the error of trend analysis, and improves the accuracy of rehabilitation deviation analysis.

[0010] Preferably, step S1 comprises the following steps:

[0011] Step S11: collecting ECG signals of the patient in different motion states by using a multi-lead electrocardiograph and a motion monitoring device to obtain ECG signals in different motion states;

[0012] Step S12: performing noise suppression processing on the ECG signals under different motion states to obtain ECG denoised signals under different motion states;

[0013] Step S13: performing time-series segmentation processing on the ECG denoising signals under different motion states to obtain ECG denoising time-series segmented signals;

[0014] Step S14: performing motion state fatigue ECG signal mapping on the ECG denoising signals under different motion states according to the ECG denoising time sequence segmented signals to obtain fatigue ECG mapping signals.

[0015] The present invention collects ECG signals of patients under different exercise states through a multi-lead electrocardiograph and a motion monitoring device, and can accurately obtain ECG data under different exercise intensities and activity conditions. This provides a rich multidimensional data source for subsequent analysis, so that the heart health status of patients during exercise can be fully monitored, especially under different exercise loads, and the reaction pattern, exercise capacity and fatigue level of the heart can be revealed. This step provides the original data basis for other processing steps. During the exercise monitoring process, the patient's exercise will generate a variety of noises, such as myoelectric noise, motion artifacts, etc., which seriously affect the accuracy and effectiveness of the electrocardiogram. By performing noise suppression processing on the collected ECG signals, unnecessary interference signals can be removed, thereby improving the quality of the ECG signals. This processing step helps to ensure that the data relied on by subsequent analysis is clearer and more accurate, and provides a reliable basis for further ECG mapping and heart rate change analysis. In the signal after noise suppression, multiple different stages of exercise states are often included, and the ECG signals of each stage have different characteristics. Through time segmentation processing, the ECG signals can be finely divided according to different exercise stages, so that the ECG signals in each exercise state can be analyzed independently. This processing step helps to capture subtle differences in different motion states, and thus provides accurate timing signal support for subsequent fatigue ECG mapping, improving the accuracy and reliability of the analysis results. Based on the ECG denoising timing segmentation signal, fatigue ECG signal mapping is performed on the signals in different motion states to obtain fatigue ECG mapping signals. This step accurately captures how the heart changes as fatigue accumulates during exercise by combining the ECG signal characteristics under motion and the physiological response to fatigue. This provides a strong basis for subsequent fatigue analysis, heart rate change monitoring, and rehabilitation deviation assessment, making fatigue assessment more refined and personalized, and providing scientific support for cardiac rehabilitation management.

[0016] Preferably, step S2 comprises the following steps:

[0017] Step S21: analyzing the heart rate variation of the fatigue ECG mapping signal to obtain fatigue heart rate variation data;

[0018] Step S22: performing waveform time domain decomposition on the fatigue heart rate variation data to obtain fatigue heart rate waveform time domain data;

[0019] Step S23: performing disordered scattered point trend quantification on fatigue heart rate change data according to fatigue heart rate waveform time domain data to obtain disordered scattered point trend quantification data of heart rate;

[0020] Step S24: performing inflection point dynamic distribution fitting on the disordered scattered trend quantified data of the heart rate to obtain trend dynamic inflection point distribution fitting data.

[0021] The present invention performs heart rate change analysis on fatigue electrocardiogram mapping signals, and can effectively extract the changing rules of the patient's heart rate during exercise. This step provides key data for evaluating the heart load and fatigue response under exercise by monitoring and analyzing the dynamic changes of heart rate. Fatigue heart rate change data can not only reflect the physiological response of the heart during exercise, but also provide intuitive basis for the subsequent fatigue assessment and the formulation of rehabilitation programs. This analysis lays an important foundation for subsequent quantitative analysis and trend identification. By performing waveform time domain decomposition on fatigue heart rate change data, the complex heart rate change signal can be finely disassembled according to the time series, and different waveform features in the signal can be identified. This process helps to distinguish various interference factors and true cardiac physiological reactions during exercise, thereby improving the resolution accuracy of the signal. Through time domain decomposition, the gradually changing heart rate waveform characteristics during fatigue can be more accurately captured, providing more detailed data support for further trend analysis and quantitative processing. According to the fatigue heart rate waveform time domain data, the disordered scattered point trend quantification of the fatigue heart rate change data can be quantified through a mathematical model. During exercise, changes in heart rate often show a certain degree of randomness or disorder, reflecting the heart's adaptation to the exercise load and signs of excessive fatigue. Through this step, the disordered trend in heart rate changes can be converted into quantifiable data, which provides a more accurate quantitative analysis tool for further fatigue assessment, anomaly detection and heart health monitoring. Fitting the dynamic distribution of inflection points of the quantitative data of the disordered scattered trend of heart rate can reveal the key turning points and dynamic change characteristics in the process of heart rate changes. This step can accurately identify the key inflection points of heart rate changes at different stages of exercise by fitting and analyzing the trend data, reflecting physiological phenomena such as increased fatigue, recovery stage or change in exercise intensity. This is crucial for monitoring the patient's physiological responses and fatigue levels at different stages of exercise, helping to detect potential problems of the heart during exercise in advance and provide a basis for accurate rehabilitation guidance.

[0022] Preferably, step S23 includes the following steps:

[0023] Step S231: performing region segmentation processing on the fatigue heart rate waveform time domain data to obtain heart rate time domain region segmentation data;

[0024] Step S232: performing extreme point distribution multi-dimensional scattered point matching on the fatigue heart rate waveform time domain data according to the heart rate time domain region segmentation data to obtain extreme point distribution multi-dimensional scattered point data;

[0025] Step S233: performing disordered scatter point regression analysis on the multi-dimensional scatter point data of extreme point distribution to obtain disordered scatter point trend regression data of heart rate;

[0026] Step S234: performing disordered scatter trend quantification on fatigue heart rate variation data according to the disordered scatter trend regression data of heart rate to obtain disordered scatter trend quantification data of heart rate.

[0027] The present invention performs regional segmentation processing on fatigue heart rate waveform time domain data, and can divide the complex heart rate waveform into different regions, each region representing a different exercise stage or heart rate response characteristics. This segmentation process can help better identify the heart rate change law in each stage, especially in the transition period or high load stage during the fatigue process. Through this method, the data of different regions can be analyzed separately, thereby revealing the relationship between heart rate change and exercise load and fatigue perception, and providing a clear basis for further quantitative analysis and trend prediction. According to the heart rate time domain regional segmentation data, multi-dimensional scatter point matching of extreme point distribution is performed, and the key extreme points in the heart rate waveform (such as the peak and valley of the heart rate) can be accurately identified, and these points are subjected to multi-dimensional scatter point matching analysis. This process not only helps to extract the key turning points of heart rate changes during exercise, but also can deeply understand the characteristics of heart rate fluctuations through multi-dimensional scatter point analysis. Through the matching of extreme point distribution, the physiological response in the fatigue process can be better revealed, especially at the critical moment of increased exercise load or gradual fatigue, providing important information for further trend analysis. The disordered scatter regression analysis of the multidimensional scatter data of the extreme point distribution can evaluate the disorder and random fluctuations in the heart rate changes through statistical methods. During exercise, the fluctuation of heart rate is often affected by many factors, such as physical exertion, fatigue accumulation, environmental conditions, etc., which lead to the disordered scatter trend of heart rate. Through regression analysis, this disordered trend can be quantified and the unpredictable fluctuation characteristics during exercise can be extracted. This analysis provides an accurate quantitative tool for further fatigue assessment and heart rate change prediction, which helps to capture the reaction law of the heart in a dynamically changing physiological environment. According to the heart rate disordered scatter trend regression data, the disordered scatter trend is quantified, and the scatter trend can be converted into specific quantitative data, thereby revealing the degree of disorder and irregular fluctuations in heart rate changes. This quantification process can help identify whether the changes in heart rate show regularity or tend to be disordered during fatigue, and thus provide a more accurate judgment of the health status of the heart. The quantified data not only helps to evaluate the physiological response during fatigue, but also provides data support for personalized exercise management programs and rehabilitation plans, and helps to formulate more scientific intervention strategies.

[0028] Preferably, step S233 includes the following steps:

[0029] Perform cluster feature extraction processing on the multi-dimensional scattered point data of extreme point distribution to obtain extreme point cluster feature data;

[0030] Perform cluster variance calculation on extreme point cluster feature data to obtain extreme point cluster variance data;

[0031] According to the extreme point clustering variance data, disordered scatter point regression analysis is performed on the extreme point distribution multidimensional scatter point data to obtain the heart rate disordered scatter point trend regression data.

[0032] The present invention performs cluster feature extraction processing on the multi-dimensional scattered point data of the extreme point distribution, and can group the key extreme points (such as peaks and valleys) in the heart rate waveform according to their characteristics, and identify the potential patterns or laws therein. Cluster processing can effectively reveal the intrinsic characteristics of heart rate fluctuations and the characteristics of heart rate changes at different stages during exercise by clustering similar extreme points together. For example, the heart rate extremes under different exercise loads show different clustering characteristics. By extracting these characteristics, a more accurate perspective can be provided for subsequent analysis, providing support for fatigue assessment and heart rate prediction. Cluster variance calculation is performed on the extreme point cluster feature data to evaluate the discrete degree or variation range of the data points in each cluster. The cluster variance can reveal the fluctuation range and stability of the heart rate extreme points in different time periods, reflecting whether the physiological response of the heart is consistent in the face of different exercise stages. A larger variance usually indicates that the heart rate fluctuation is larger, which is related to increased fatigue or changes in exercise intensity, while a smaller variance indicates that the heart rate change tends to be stable. Therefore, the calculation of cluster variance provides a quantitative basis for analyzing the trend of heart rate fluctuations, which can help further judge the exercise load and fatigue state. The randomness and disorder in heart rate fluctuations can be evaluated by performing unordered scatter point regression analysis on the multidimensional scatter point data of extreme point distribution based on the extreme point clustering variance data. Through regression analysis, the unordered scatter point trend in the heart rate data can be transformed into a quantitative regression model to reveal the nonlinear characteristics of heart rate changes. This step is particularly helpful in identifying irregular fluctuations in heart rate during exercise due to load changes, increased fatigue, or cardiac adaptation responses. This regression analysis can not only help establish a mathematical model of heart rate changes, but also provide a reliable quantitative tool for fatigue monitoring and rehabilitation guidance, which helps to more accurately predict and intervene in individual cardiac responses.

[0033] Preferably, step S24 includes the following steps:

[0034] Step S241: performing trend rising boundary segmentation processing on the disordered scattered point trend quantization data of the heart rate to obtain the heart rate trend rising boundary segmentation data;

[0035] Step S242: performing inflection point distribution space identification on the disordered scattered point trend quantization data of the heart rate according to the heart rate trend rising boundary segmentation data to obtain inflection point distribution space identification data;

[0036] Step S243: Calculate the distribution time series difference of the inflection point distribution space identification data to obtain the inflection point distribution space time series difference;

[0037] Step S244: performing inflection point dynamic distribution fitting on the disordered scattered trend quantification data of the heart rate according to the inflection point distribution spatial time series difference to obtain trend dynamic inflection point distribution fitting data.

[0038] The present invention performs trend rising boundary segmentation processing on disordered scattered trend quantization data of heart rate, and can identify the obvious rising stage and its boundary in the process of heart rate change, so as to separate the different stages in the process of heart rate change. This segmentation processing helps to accurately capture the rising trend in heart rate fluctuation, especially when the exercise load gradually increases or fatigue gradually increases, the rising law of heart rate. By identifying the rising boundary, the analysis can more clearly reveal the critical point of heart rate rise, help judge the physiological changes and fatigue process of individuals during exercise, and provide clear data support for subsequent trend analysis and fatigue prediction. According to the heart rate trend rising boundary segmentation data, the inflection point distribution space of the disordered scattered trend quantization data of heart rate is identified, and the key inflection point in the heart rate waveform can be accurately identified, that is, the turning point from rising to stable, or from stable to falling. Through this spatial identification, the dynamic characteristics in the process of heart rate change can be analyzed, revealing the speed and amplitude changes of heart rate fluctuations at different time points. The inflection point distribution space identification helps to accurately capture the key change points in the heart rate waveform, and provides an important data basis for subsequent timing difference calculation and dynamic fitting, so that the analysis results are more accurate. The distribution time series difference calculation of the inflection point distribution space identification data aims to quantify the time difference between the inflection points and the relative time order of each inflection point. This calculation process helps analyze the time series relationship between different inflection points and reveals the periodicity or disorder of heart rate changes. Through the calculation of time series difference, the time delay of the inflection point in the heart rate fluctuation can be identified, and then the fatigue accumulation or dynamic changes of physical recovery during exercise can be evaluated. The calculation of time series difference can not only reflect the rhythm of heart rate changes, but also help detect unstable or nonlinear heart rate change patterns during exercise, providing data support for accurate monitoring of exercise status. According to the time series difference of the inflection point distribution space, the dynamic distribution of the inflection point is fitted to the quantitative data of the disordered scattered trend of the heart rate. Based on the previous analysis results, a dynamic mathematical model can be established to describe the distribution law of the inflection points in the process of heart rate changes. Through dynamic fitting, the long-term trend and short-term fluctuation of heart rate fluctuations during exercise can be captured, thereby reflecting the adaptability and fatigue state of the heart under different loads. The fitting results can not only provide accurate parameters for the prediction of heart rate changes, but also provide quantitative basis for personalized exercise programs, health management and fatigue monitoring. This process helps to conduct in-depth analysis of the dynamic changes in heart rate and provide scientific guidance for the training and recovery of athletes and individuals.

[0039] Preferably, step S3 comprises the following steps:

[0040] Step S31: performing multi-scale rule intersection analysis on fatigue heart rate change data according to trend dynamic inflection point distribution fitting data to obtain fatigue heart rate change rule intersection data;

[0041] Step S32: performing resting heart rate difference evaluation on the fatigue heart rate change rule intersection data to obtain resting heart rate difference evaluation data;

[0042] Step S33: simulating the heart fatigue level according to the intersection data of fatigue heart rate change rules and the resting heart rate difference evaluation data to obtain heart fatigue level simulation data;

[0043] Step S34: Perform a quantitative assessment of rehabilitation deviation based on the cardiac fatigue level simulation data, the resting heart rate difference assessment data, and the fatigue heart rate change rule intersection data to obtain rehabilitation deviation assessment data.

[0044] The present invention performs multi-scale regular intersection analysis on fatigue heart rate change data based on trend dynamic inflection point distribution fitting data, aiming to reveal different regularities and patterns in heart rate changes from a multi-level perspective. This analysis method can identify the commonalities and differences of heart rate fatigue changes at different time scales, and help determine the key turning points and regular characteristics of heart rate changes. Through intersection analysis, multiple data levels can be more comprehensively integrated, the heart rate fatigue assessment model can be optimized, and a more accurate basis can be provided for subsequent fatigue monitoring and personalized intervention. The resting heart rate difference assessment is performed on the intersection data of fatigue heart rate change rules, mainly by comparing the heart rate difference between individuals in fatigue state and resting state, to assess the body's recovery ability and adaptability under different fatigue levels. Resting heart rate is an important indicator for assessing heart health and fatigue level. A resting heart rate with a large difference reflects excessive load or insufficient recovery of the heart. Through this assessment step, a scientific basis can be provided for the accurate quantification of fatigue level, and data support can be provided for individual recovery plans. According to the intersection data of fatigue heart rate change rules and the resting heart rate difference assessment data, the heart fatigue level simulation is performed, aiming to quantify different levels of heart fatigue by establishing a mathematical model. The simulation combines multi-dimensional heart rate data to objectively evaluate the individual's cardiac fatigue level and classify fatigue into different levels, such as mild, moderate and severe. The simulation of cardiac fatigue level provides a basis for individuals to formulate appropriate exercise loads, rest time and recovery plans, which helps to avoid cardiac health problems caused by overtraining and improve exercise effects and recovery efficiency. Based on the simulation data of cardiac fatigue level, the resting heart rate difference assessment data and the intersection data of fatigue heart rate change rules, the quantitative assessment of rehabilitation deviation can quantitatively assess the degree of deviation of individuals in the rehabilitation process, that is, the difference between the actual rehabilitation state and the ideal recovery state. Through this assessment, potential problems in the rehabilitation process of individuals can be identified, such as excessive fatigue, insufficient recovery or cardiac adaptation problems during the recovery process. The quantitative assessment of rehabilitation deviation provides a scientific reference for adjusting rehabilitation plans, personalized health management and avoiding sports injuries, helping individuals achieve the best recovery effect during the rehabilitation process.

[0045] Preferably, step S33 includes the following steps:

[0046] Step S331: extracting the intersection feature vector of the heart rate change law from the intersection data of the fatigue heart rate change law to obtain the intersection vector of the fatigue heart rate law;

[0047] Step S332: performing multimodal fusion processing on the resting heart rate difference evaluation data according to the fatigue heart rate rule intersection vector to obtain fatigue heart rate multimodal fusion data;

[0048] Step S333: Perform heart fatigue level simulation based on fatigue heart rate multimodal fusion data to obtain heart fatigue level simulation data.

[0049] The present invention can extract key regular features from complex heart rate data and form a fatigue heart rate law intersection vector by extracting feature vectors from the intersection data of fatigue heart rate changes. This feature vector can effectively express the change pattern of heart rate in different time periods and different exercise intensities, and provide highly compressed and meaningful data representation for subsequent analysis. By extracting these feature vectors, the trend of individual heart rate fatigue can be better captured, providing data support for cardiac fatigue assessment, thereby improving the judgment accuracy of fatigue degree and the operability of analysis. By performing multimodal fusion processing on the intersection vector of fatigue heart rate law and the resting heart rate difference assessment data, data from multiple different sources can be integrated, thereby improving the comprehensiveness and accuracy of analysis. The resting heart rate difference assessment reflects the individual's recovery state, while the intersection vector of fatigue heart rate law reveals the heart rate change pattern during exercise. By fusing these two data sources, the comprehensive characteristics of the individual heart under different physiological states can be captured, making the assessment of fatigue state more comprehensive, and providing more accurate data support for further fatigue analysis. Based on the fatigue heart rate multimodal fusion data, cardiac fatigue level simulation is performed, aiming to simulate and evaluate the individual's cardiac fatigue level by comprehensively analyzing all relevant data. This step converts multimodal fusion data into different levels of cardiac fatigue, such as mild, moderate or severe fatigue, by constructing a mathematical model or machine learning algorithm. Through the simulation of cardiac fatigue levels, more refined fatigue monitoring can be achieved, helping individuals or athletes make corresponding adjustments at different exercise loads and recovery stages to improve training effects and reduce the risk of overtraining. This simulation not only enhances the accuracy of fatigue assessment, but also provides an important decision-making basis for personalized training and rehabilitation plans.

[0050] Preferably, step S34 includes the following steps:

[0051] Step S341: performing differential feature extraction on the resting heart rate difference assessment data according to the heart fatigue level simulation data to obtain heart rate differential feature data;

[0052] Step S342: performing polynomial fitting analysis on the heart rate difference feature data and the intersection data of fatigue heart rate variation law to obtain the fatigue heart rate polynomial fitting coefficient;

[0053] Step S343: performing nonlinear mapping transformation on the heart rate difference characteristic data according to the fatigue heart rate polynomial fitting coefficient to obtain the heart rate nonlinear characteristic mapping data;

[0054] Step S344: Perform quantitative assessment of rehabilitation deviation based on the heart rate nonlinear characteristic mapping data, fatigue heart rate polynomial fitting coefficients and heart rate differential characteristic data to obtain rehabilitation deviation assessment data.

[0055] The present invention extracts differential features from resting heart rate difference assessment data based on heart fatigue level simulation data, with the purpose of extracting important dynamic features reflecting heart recovery ability and load changes by calculating the changing trend of resting heart rate difference. The differential feature can effectively capture the rate and trend of change of heart rate, thereby revealing the fatigue recovery of individuals at different stages. This process provides accurate basic data for subsequent rehabilitation assessment by quantifying the changes in heart rate differences, helping to identify potential problems in rehabilitation progress. By performing polynomial fitting analysis on the intersection data of heart rate differential feature data and fatigue heart rate change law, the purpose is to establish a mathematical relationship model between the two. This process can accurately reflect the complex laws of heart rate changes, including linear and nonlinear change trends, by fitting a high-order polynomial model. The polynomial fitting coefficient not only helps to reveal the deep-level laws in the fatigue and recovery process, but also provides quantifiable parameter support for subsequent nonlinear mapping and deviation evaluation, thereby improving the accuracy and reliability of fatigue evaluation. According to the fatigue heart rate polynomial fitting coefficient, the heart rate differential feature data is nonlinearly mapped and transformed. This transformation can map the original linear data into a nonlinear space that is more in line with actual physiological changes through mathematical transformation, so as to better adapt to the complex dynamic characteristics of heart rate data. The application of nonlinear mapping can capture some subtle patterns of heart rate changes, such as nonlinear fatigue accumulation and heterogeneity in the recovery process, which helps to improve the accuracy of rehabilitation deviation assessment and optimize individual recovery strategies. The quantitative assessment of rehabilitation deviation is based on heart rate nonlinear feature mapping data, fatigue heart rate polynomial fitting coefficients and heart rate differential feature data, with the aim of comprehensively assessing the individual's recovery progress during the rehabilitation process and its deviation from the ideal recovery state. By combining multiple data sources and nonlinear mapping results, the quantitative assessment of rehabilitation deviation can more accurately reveal problems in the rehabilitation process, such as insufficient recovery, excessive fatigue or mismatched recovery cycles. This assessment method provides data support for personalized rehabilitation plans and helps to formulate more appropriate exercise intensity and rest strategies to promote optimal recovery effects.

[0056] Preferably, the present invention further provides an ECG monitoring system based on cardiac rehabilitation data, which is used to perform the ECG monitoring method based on cardiac rehabilitation data as described above. The ECG monitoring system based on cardiac rehabilitation data comprises:

[0057] The ECG signal mapping module is used to collect ECG signals of patients in different motion states through a multi-lead ECG instrument and a motion monitoring device to obtain ECG signals in different motion states; perform motion state fatigue ECG signal mapping on the ECG signals in different motion states to obtain fatigue ECG mapping signals;

[0058] The inflection point dynamic distribution fitting module is used to analyze the heart rate change of the fatigue ECG mapping signal to obtain fatigue heart rate change data; to quantify the disordered scattered point trend of the fatigue heart rate change data to obtain the disordered scattered point trend quantification data of the heart rate; to perform inflection point dynamic distribution fitting on the disordered scattered point trend quantification data of the heart rate to obtain the trend dynamic inflection point distribution fitting data;

[0059] The rehabilitation deviation quantitative assessment module is used to perform multi-scale rule intersection analysis on fatigue heart rate change data according to the trend dynamic inflection point distribution fitting data to obtain fatigue heart rate change rule intersection data; perform rehabilitation deviation quantitative assessment based on fatigue heart rate change rule intersection data to obtain rehabilitation deviation assessment data;

[0060] The terminal feedback module is used to send the rehabilitation deviation assessment data to the terminal to perform ECG monitoring of cardiac rehabilitation data.

[0061] The beneficial effect of the present invention is that the ECG signals of patients in different exercise states are collected by a multi-lead electrocardiograph and a motion monitoring device, and the ECG data during exercise can be fully obtained. This process can not only reflect the impact of different exercise intensities on the heart, but also extract the signal characteristics related to fatigue by mapping the fatigue degree of the motion state of the ECG signal, and further provide accurate basic data for subsequent analysis and evaluation. This lays a solid foundation for accurately judging the patient's exercise load and fatigue level, and is helpful for the formulation of personalized rehabilitation programs. Heart rate change analysis of fatigue ECG mapping signals can reveal the patient's heart rate fluctuations during exercise, and then reflect the heart's response and adaptability to exercise load. Through the quantification of disordered scatter trend, the disordered characteristics in the heart rate change can be extracted, thereby quantifying the stress response and recovery state of the heart. Then, the dynamic distribution fitting technology of inflection points can be applied to further reveal the dynamic inflection points and turning points that occur during the heart rate change process, which helps to deeply understand the abnormal changes in the exercise process, and then achieve accurate heart health monitoring. According to the trend dynamic inflection point distribution fitting data, multi-scale rule intersection analysis can be performed to identify the inherent rules of fatigue heart rate changes and reveal the complexity and diversity of heart rate changes under different exercise loads. Through this analysis, the key rules of fatigue heart rate changes can be extracted to provide a scientific basis for further rehabilitation evaluation. Based on the intersection data of fatigue heart rate changes, quantitative evaluation of rehabilitation deviation can not only quantify the deviation in the rehabilitation process, but also provide accurate reference for the formulation of personalized rehabilitation plans, ensuring that patients receive more scientific and accurate rehabilitation guidance. By sending the rehabilitation deviation evaluation data to the terminal, the patient's cardiac rehabilitation status can be monitored in real time. The implementation of this process ensures dynamic monitoring and feedback of heart health during the rehabilitation process, and provides real-time data support for the patient's exercise intensity and recovery. In addition, the feedback function of the terminal can help doctors and patients instantly understand the progress and deviation of cardiac rehabilitation, and then make timely adjustments during the rehabilitation process, optimize cardiac rehabilitation strategies, improve rehabilitation effects, and reduce the risk of recurrence of heart disease. Therefore, the present invention optimizes a traditional ECG monitoring method based on cardiac rehabilitation data, solves the problems of large trend analysis error and inaccurate rehabilitation deviation analysis in the traditional ECG monitoring method based on cardiac rehabilitation data, reduces the error of trend analysis, and improves the accuracy of rehabilitation deviation analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 A schematic diagram of the steps of an electrocardiogram monitoring method based on cardiac rehabilitation data;

[0063] Figure 2 for Figure 1 Detailed implementation steps of step S2 in the flowchart;

[0064] Figure 3 for Figure 1 Detailed implementation steps of step S3 in FIG.

[0065] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0066] The technical method of the present invention is described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0067] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0068] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0069] To achieve this, please refer to Figures 1 to 3 , an electrocardiogram monitoring method based on cardiac rehabilitation data, the method comprising the following steps:

[0070] Step S1: collecting ECG signals of the patient in different motion states by means of a multi-lead electrocardiograph and a motion monitoring device to obtain ECG signals in different motion states; performing motion state fatigue ECG signal mapping on the ECG signals in different motion states to obtain fatigue ECG mapping signals;

[0071] Step S2: analyzing the heart rate variation of the fatigue ECG mapping signal to obtain fatigue heart rate variation data; performing disordered scattered point trend quantification on the fatigue heart rate variation data to obtain disordered scattered point trend quantification data of the heart rate; performing inflection point dynamic distribution fitting on the disordered scattered point trend quantification data of the heart rate to obtain trend dynamic inflection point distribution fitting data;

[0072] Step S3: performing multi-scale rule intersection analysis on fatigue heart rate change data according to trend dynamic inflection point distribution fitting data to obtain fatigue heart rate change rule intersection data; performing quantitative assessment of rehabilitation deviation based on fatigue heart rate change rule intersection data to obtain rehabilitation deviation assessment data;

[0073] Step S4: Send the rehabilitation deviation assessment data to the terminal to perform ECG monitoring of cardiac rehabilitation data.

[0074] In the embodiment of the present invention, reference Figure 1 FIG. 1 is a schematic diagram of a step flow chart of an electrocardiogram monitoring method based on cardiac rehabilitation data of the present invention. In this example, the electrocardiogram monitoring method based on cardiac rehabilitation data includes the following steps:

[0075] Step S1: collecting ECG signals of the patient in different motion states by means of a multi-lead electrocardiograph and a motion monitoring device to obtain ECG signals in different motion states; performing motion state fatigue ECG signal mapping on the ECG signals in different motion states to obtain fatigue ECG mapping signals;

[0076] @First, the patient's ECG signals under different motion states are collected through a multi-lead electrocardiograph (such as a 12-lead electrocardiograph) and a motion monitoring device (such as a portable motion sensor or a smart watch). The multi-lead sensor of the electrocardiograph records the ECG signals in real time. The data collected by the device includes the potential changes of each electrode contact, and the generated ECG signals are transmitted to the central processing system via wired or wireless means. The motion monitoring device monitors the motion state in real time through accelerometers, gyroscopes, etc., and obtains data such as gait, motion intensity, motion frequency, and posture changes. Combining the data of these two types of devices, the patient's motion state under different motion intensities (for example, walking, jogging, cycling, and stationary states) can be determined. Subsequently, the collected ECG signals are mapped to the degree of fatigue of the motion state. The mapping method is based on multimodal data fusion. By associating the characteristics of the ECG signals with the motion state data (such as acceleration and motion frequency), the patient's motion fatigue is obtained. This process uses signal processing technology to filter the ECG signals under different motion states, remove interference signals and noise, extract frequency domain features in the ECG through Fourier transform and other algorithms, and combine the features in the motion data (such as changes in motion intensity) to form a fatigue ECG mapping signal. The fatigue ECG signal obtained by processing the mapping signal contains the relationship between the changes in the ECG signal during exercise and the degree of fatigue, which can then be used for further analysis in subsequent steps.

[0077] Step S2: analyzing the heart rate variation of the fatigue ECG mapping signal to obtain fatigue heart rate variation data; performing disordered scattered point trend quantification on the fatigue heart rate variation data to obtain disordered scattered point trend quantification data of the heart rate; performing inflection point dynamic distribution fitting on the disordered scattered point trend quantification data of the heart rate to obtain trend dynamic inflection point distribution fitting data;

[0078] @First, the heart rate variation analysis of fatigue ECG mapping signals is carried out to calculate the heart rate variation of patients under different exercise states. The peak detection algorithm (for example, the R wave detection method based on dynamic threshold) is used to extract the R wave peak from the ECG signal, and the heart rate variation data is obtained by calculating the time interval between R waves. The heart rate data includes information such as resting heart rate, heart rate changes in different time periods during exercise, and the rate of heart rate acceleration and deceleration. Next, the heart rate variation data is quantified by disordered scattered point trend. First, the heart rate variation data is analyzed in the time domain to identify the peaks and valleys of each heart rate fluctuation in the data. Then, the heart rate variation data is converted into disordered scattered points by using statistical methods, which avoids the interference of time series features in the data. On this basis, the disordered scattered point data is normalized by using the trend quantification algorithm to obtain the disordered scattered point trend quantification data of heart rate. This process helps to reveal the trend of heart rate changes during fatigue and avoids the influence of noise and irregular data. Finally, the inflection point dynamic distribution fitting is performed on the disordered scattered point trend quantification data of heart rate. By analyzing the turning points of the heart rate trend (such as the points where the heart rate suddenly rises or falls), the scattered data is dynamically distributed and fitted using curve fitting methods (such as the least squares method) to obtain the trend dynamic turning point distribution fitting data. This data describes the distribution characteristics of each important turning point in the heart rate change process and can reveal the different fatigue stages of patients during exercise.

[0079] Step S3: performing multi-scale rule intersection analysis on fatigue heart rate change data according to trend dynamic inflection point distribution fitting data to obtain fatigue heart rate change rule intersection data; performing quantitative assessment of rehabilitation deviation based on fatigue heart rate change rule intersection data to obtain rehabilitation deviation assessment data;

[0080] @According to the trend dynamic inflection point distribution fitting data, a multi-scale law intersection analysis of fatigue heart rate change data is performed. This analysis reveals the multi-level laws hidden in the data by decomposing the data at different scales. For example, wavelet transform is used to decompose the data into low-frequency parts (representing long-term trends) and high-frequency parts (representing short-term fluctuations), and then the laws at different scales are compared through intersection analysis to obtain the intersection data of fatigue heart rate change laws. This data contains the heart rate change laws of patients under different exercise states, revealing the degree of fatigue accumulation during the patient's exercise. Subsequently, a quantitative assessment of rehabilitation deviation is performed based on the intersection data of fatigue heart rate change laws. This assessment process calculates the deviation between the patient's heart rate change and the expected rehabilitation trajectory by comparing it with the normal heart rate change law. First, the reference heart rate change model (for example, the heart rate change trajectory of a healthy individual) is used to compare with the patient's heart rate change law to calculate the degree of difference; second, the size of the rehabilitation deviation is quantitatively assessed according to the degree of difference. This data reflects the difference between the patient's ideal state during the rehabilitation process and helps doctors evaluate the rehabilitation progress.

[0081] Step S4: sending the rehabilitation deviation assessment data to the terminal to perform ECG monitoring of cardiac rehabilitation data;

[0082] @Send the rehabilitation deviation assessment data to the terminal device for patients or medical staff to perform ECG monitoring of cardiac rehabilitation data. First, the rehabilitation deviation assessment data is transmitted to the patient's smart terminal device (such as a mobile phone or tablet) through wired or wireless communication technology (such as Wi-Fi or Bluetooth). The terminal device displays the rehabilitation assessment data through a dedicated application and provides real-time feedback. For example, the interface displays fatigue ECG mapping signals, heart rate change trends, rehabilitation deviation assessment results, etc. The terminal can also generate reports based on these data for doctors to refer to, and then adjust the patient's exercise and rehabilitation plan. This feedback mechanism ensures that patients can obtain relevant information about cardiac rehabilitation in a timely manner and make necessary adjustments to promote the optimization of the rehabilitation process.

[0083] Preferably, step S1 comprises the following steps:

[0084] Step S11: collecting ECG signals of the patient in different motion states by using a multi-lead electrocardiograph and a motion monitoring device to obtain ECG signals in different motion states;

[0085] Step S12: performing noise suppression processing on the ECG signals under different motion states to obtain ECG denoised signals under different motion states;

[0086] Step S13: performing time-series segmentation processing on the ECG denoising signals under different motion states to obtain ECG denoising time-series segmented signals;

[0087] Step S14: performing motion state fatigue ECG signal mapping on the ECG denoising signals under different motion states according to the ECG denoising time sequence segmented signals to obtain fatigue ECG mapping signals.

[0088] @By combining a multi-lead electrocardiograph and a motion monitoring device, the patient's ECG signals under different motion states are collected. The multi-lead electrocardiograph detects ECG signals through its multiple electrode sensors (such as 12-lead electrodes). When each electrode contacts the patient's skin surface, it records the electrophysiological signals and converts them into potential changes, eventually forming a complete ECG signal. These signals reflect each cycle of the heart's electrical activity and are transmitted to the central processing unit through a computer system. In parallel, motion monitoring devices, such as portable accelerometers, gyroscopes, or sports bracelets, monitor the patient's motion state in real time. These devices can collect data such as gait, motion frequency, and motion intensity. The motion monitoring device continuously records the changes in motion (such as walking, running, and stillness) and transmits these data synchronously to the same processing unit. By synchronously collecting ECG signals and motion state signals, the relationship between the patient's ECG signals and motion load under different motion states can be obtained, providing a basis for subsequent signal processing and analysis. Remove ECG signal noise caused by environmental interference (such as myoelectric noise, motion artifacts, etc.). First, the collected ECG signal will be affected by factors such as muscle activity and electrical equipment during exercise, resulting in noise components in the signal. In this step, filtering technology is applied to suppress the noise of the signal. Specifically, a bandpass filter is used to set a frequency range (for example, 0.5 Hz to 50 Hz), which includes the main frequency components of the ECG signal, while excluding low-frequency noise (such as myoelectric noise) and high-frequency noise (such as electromagnetic interference). The filtered signal still contains some high-frequency components or artifacts. At this time, wavelet transform (such as Daubechies wavelet transform) can be further used to decompose and extract the low-frequency part of the ECG signal. For the denoised signal, adaptive filtering technology, such as LMS (least mean square) adaptive algorithm, can be applied to adjust the filtering parameters according to real-time environmental changes to further enhance the denoising effect. The filtered ECG signal is relatively clean and can more accurately reflect the patient's cardiac electrical activity. The denoised ECG signal is segmented according to time to facilitate subsequent analysis and mapping operations. In order to accurately divide the ECG signals under different motion states, it is first necessary to determine the time interval in the signal. For example, during the acquisition process, the signal can be divided into a motion period and a stationary period. According to the motion state data obtained by the motion monitoring device, the ECG signal can be divided between the motion period and the stationary period. By setting a time window (for example, every 10 seconds or 30 seconds as a segment), the continuous ECG signal is divided into multiple time segments. The data of each time segment contains the ECG changes within the corresponding time window, which can reveal the patient's cardiac electrical activity in different time periods. In the process of signal time segmentation, the slope threshold algorithm can also be used to detect drastic changes in the ECG signal, such as a sudden acceleration of the heartbeat or abnormal heart rhythm fluctuations, so as to further refine the signal segment. The time label and signal characteristics (such as heart rate, waveform, etc.) of each signal segment will be used as the basis for analysis in the subsequent mapping process.Based on the ECG denoising time-series segmented signal, the ECG signal mapping of the fatigue degree of the exercise state is performed. The purpose of this mapping process is to analyze and quantify the patient's fatigue degree according to the changes in the ECG signal under different exercise states. First, the ECG signal of each time segment is mapped to the corresponding fatigue signal in combination with the exercise state data (such as exercise intensity, exercise time, heart rate change, etc.) provided by the exercise monitoring device. In order to achieve this mapping, a mapping function based on empirical rules can be used to define the relationship between the waveform characteristics (such as R wave amplitude, QT interval, etc.) in the ECG signal and the exercise load (such as step frequency, exercise speed). Through regression analysis (such as linear regression or polynomial regression), the exercise state data and the ECG data are combined to obtain the fatigue signal. This process can also use the cluster analysis method to group the ECG signals under different exercise states, determine the ECG characteristics under different exercise loads, and then map them to the corresponding fatigue level. The mapping signal reflects the relationship between the ECG signal changes and fatigue of the patient during exercise. The generated fatigue ECG mapping signal can be used for subsequent heart rate change analysis and rehabilitation evaluation to ensure effective monitoring and intervention during the rehabilitation process.

[0089] Preferably, step S2 comprises the following steps:

[0090] Step S21: analyzing the heart rate variation of the fatigue ECG mapping signal to obtain fatigue heart rate variation data;

[0091] Step S22: performing waveform time domain decomposition on the fatigue heart rate variation data to obtain fatigue heart rate waveform time domain data;

[0092] Step S23: performing disordered scattered point trend quantification on fatigue heart rate change data according to fatigue heart rate waveform time domain data to obtain disordered scattered point trend quantification data of heart rate;

[0093] Step S24: performing inflection point dynamic distribution fitting on the disordered scattered trend quantified data of the heart rate to obtain trend dynamic inflection point distribution fitting data.

[0094] As an example of the present invention, refer to Figure 2 As shown, in this example, step S2 includes:

[0095] Step S21: analyzing the heart rate variation of the fatigue ECG mapping signal to obtain fatigue heart rate variation data;

[0096] @The goal of step S21 is to analyze the heart rate changes in the fatigue ECG mapping signal and obtain fatigue heart rate change data. First, extract the RR interval in the ECG signal, that is, the time interval between two adjacent R waves, which reflects the heart's beating cycle. By calculating the changes in the RR interval in each time period, a heart rate fluctuation sequence is obtained. Specifically, the RR interval (unit: second) can be automatically calculated by the waveform recognition algorithm of the ECG signal. The heart rate data is obtained by taking the reciprocal of the RR interval (unit: bpm, that is, the number of beats per minute). When analyzing the fatigue ECG mapping signal, it is necessary to count the heart rate changes in different time periods. For example, a sliding window algorithm can be used to divide the signal into multiple time windows (for example, every 10 seconds or 30 seconds as a window), calculate the heart rate average in each window, and record the maximum fluctuation range of the heart rate. Through this analysis, the heart rate change data reflecting the fatigue process are obtained. These data can reveal the changes in the patient's heart burden under different exercise states, and thus provide a basis for subsequent fatigue analysis.

[0097] Step S22: performing waveform time domain decomposition on the fatigue heart rate variation data to obtain fatigue heart rate waveform time domain data;

[0098] @The task of step S22 is to perform waveform time domain decomposition on fatigue heart rate change data in order to further extract the time domain characteristics of heart rate. First, fatigue heart rate change data is decomposed into multiple components of different frequency bands as a time series. This process is achieved through time domain analysis methods. Common time domain decomposition methods include local weighted regression, Fourier transform or wavelet decomposition. Specifically, wavelet transform (such as Daubechies wavelet) can be applied to perform multi-scale decomposition on heart rate change data. Wavelet transform can decompose heart rate data into different frequency components, so as to analyze short-term fluctuations (such as fast rhythm changes) and long-term changes (such as slow fatigue process) respectively. Each decomposed signal represents a different frequency band of the heart rate waveform, which can reveal the instantaneous fluctuations and overall change trends of heart rate during exercise. Each decomposed frequency band will reflect heart rate fluctuations at different levels. For example, low-frequency components represent long-term fatigue processes, while high-frequency components reveal short-term fluctuations or stress responses. Through these decomposed time domain data, we can have a deeper understanding of the change mechanism of heart rate under fatigue state and provide a basis for subsequent trend quantification.

[0099] Step S23: performing disordered scattered point trend quantification on fatigue heart rate change data according to fatigue heart rate waveform time domain data to obtain disordered scattered point trend quantification data of heart rate;

[0100] @Quantify the disordered scattered point trend of fatigue heart rate waveform time domain data. The purpose of this process is to quantify the disorder and trend change of heart rate fluctuations through disordered scattered point analysis technology. First, the key features (such as fluctuation amplitude, waveform shape, periodicity, etc.) extracted from the fatigue heart rate waveform time domain data are converted into a scattered point data set, and each scattered point represents the heart rate change state at a specific moment. Next, the disordered scattered point analysis technology (such as scatter plot method) is used to visualize these scattered point data to identify the trend changes of heart rate fluctuations in different time periods. The specific operation steps include: 1) assigning a coordinate axis value to the heart rate data of each time window; 2) arranging the data points in the scatter plot according to the amplitude of heart rate change to form the scattered point trend of heart rate fluctuations; 3) quantifying the trend of these disordered scattered points through fitting algorithms (such as least squares method or local weighted regression method) to obtain the quantitative trend characteristics of heart rate fluctuations. This process provides a data basis for the subsequent dynamic distribution fitting of inflection points by quantifying the disordered scattered point trend of heart rate fluctuations. The disordered scattered trend quantitative data can reveal the fluctuation pattern and mutation characteristics of the patient's heart rate during exercise fatigue.

[0101] Step S24: performing inflection point dynamic distribution fitting on the disordered scattered trend quantified data of the heart rate to obtain trend dynamic inflection point distribution fitting data;

[0102] @The goal of step S24 is to fit the dynamic distribution of inflection points for the disordered scattered trend quantified data of heart rate. First, based on the scattered trend data obtained in the previous step, the inflection point detection algorithm is used to identify the inflection point position in the data change. The inflection point refers to the position where the curvature in the heart rate change curve changes suddenly, which usually reflects the key turning point in the heart rate fluctuation, such as the change in exercise intensity or the change in fatigue level. In order to accurately fit the dynamic distribution of the inflection point, the scattered trend data can be processed by a curve fitting algorithm (such as spline interpolation or least squares method). In the fitting process, the trend curve is first determined according to the scattered data, and the inflection point in the heart rate fluctuation is identified according to the curvature of the curve. Each inflection point corresponds to a specific moment and change amplitude, which can reveal the key moment of the patient's heart rate change. The result of this process is a set of dynamic distribution data describing the inflection point of heart rate fluctuation. These data can help further analyze the law of heart rate change in patients under fatigue, reveal the mutation pattern of heart rate during exercise load or recovery, and provide effective data information for subsequent ECG monitoring and rehabilitation intervention.

[0103] Preferably, step S23 includes the following steps:

[0104] Step S231: performing region segmentation processing on the fatigue heart rate waveform time domain data to obtain heart rate time domain region segmentation data;

[0105] Step S232: performing extreme point distribution multi-dimensional scattered point matching on the fatigue heart rate waveform time domain data according to the heart rate time domain region segmentation data to obtain extreme point distribution multi-dimensional scattered point data;

[0106] Step S233: performing disordered scatter point regression analysis on the multi-dimensional scatter point data of extreme point distribution to obtain disordered scatter point trend regression data of heart rate;

[0107] Step S234: performing disordered scatter trend quantification on fatigue heart rate variation data according to the disordered scatter trend regression data of heart rate to obtain disordered scatter trend quantification data of heart rate.

[0108] @Perform regional segmentation processing on the fatigue heart rate waveform time domain data so as to extract representative time periods from the overall waveform. This process first requires a detailed analysis of the fatigue heart rate waveform time domain data to identify different change intervals in the waveform. In actual operation, the peak detection algorithm (such as a threshold-based method) is first applied to identify the local maximum and minimum values ​​in the waveform. These extreme points usually mark the turning point of the change trend of the heart rate waveform. Then, the heart rate waveform is divided into different regions according to these extreme points, and each region corresponds to a different stage of heart rate fluctuation (such as the beginning of exercise, fatigue process, recovery, etc.). In specific operations, the heart rate waveform can be segmented by sliding windows. For example, the window length can be set to a fixed time period (such as 10 seconds), and each sliding advances a certain step length. In each time window, by calculating the heart rate change range and fluctuation amplitude of each window, it is determined whether the window meets a specific regional feature (such as a stable period, a rapid growth period, a gradual weakening period, etc.). Once these regions are identified, the heart rate waveform can be regionalized to obtain the heart rate time domain regional segmentation data. The goal of step S232 is to further perform multi-dimensional scattered point matching of extreme point distribution on fatigue heart rate waveform time domain data according to the heart rate time domain area segmentation data. First, the extreme points in each area are extracted through the regional division data of the heart rate waveform obtained in the previous step. These extreme points reflect the maximum amplitude and mutation point of the heart rate fluctuation. Each extreme point corresponds to an important turning point of the heart rate fluctuation. In order to perform multi-dimensional scattered point matching, these extreme points need to be marked according to different attributes (such as fluctuation amplitude, time position, relationship with previous and next extreme points, etc.). During the processing, a multi-dimensional data matching algorithm (such as a distance-based clustering algorithm or a K-means algorithm) can be used to group and match extreme points in different regions. Each extreme point will be matched with other extreme points to form a multi-dimensional scattered point set, which represents the distribution of extreme points in the heart rate waveform and their mutual relationship. The key to this process is how to choose a suitable matching standard. These points can be accurately matched according to factors such as the timing characteristics, fluctuation amplitude and position relationship of the extreme points, so as to obtain multi-dimensional scattered point data of the extreme point distribution. This data set can provide key information at different stages of heart rate fluctuations and provide effective input data for subsequent unordered scatter point regression analysis. Unordered scatter point regression analysis is performed on the multidimensional scatter point data of extreme point distribution. First, the multidimensional scatter point data of extreme point distribution provides the key turning points of heart rate fluctuations and their relative positions. On this basis, regression analysis techniques (such as polynomial regression, local weighted regression, etc.) are used to model these scattered points. The purpose of regression analysis is to establish a mathematical model to describe the relationship and trend changes between extreme points. Specifically, for each extreme point, by calculating the distance between it and the adjacent extreme point, the amplitude of fluctuation and other characteristics, a regression model is established to fit the entire scattered point data set.The least square method can be used to optimize the parameters of the regression function so that the regression curve can best fit all scattered data. The result of the regression analysis is to obtain a trend curve describing the distribution of extreme points and quantify the law of heart rate fluctuation. This process can reveal the nonlinear characteristics and complex change patterns of heart rate fluctuations, and provide valuable data information for further trend quantification. According to the disordered scattered trend regression data of heart rate, the disordered scattered trend of fatigue heart rate change data is quantified. In this process, the disordered scattered regression data obtained in step S233 is first used as a basis to further quantify the trend characteristics of heart rate fluctuations. The specific operation is to perform mathematical analysis on the regression curve to calculate parameters such as the slope, inflection point and change amplitude of the curve. These parameters can reflect the key changes in the heart rate fluctuation process, such as the intensification, weakening or periodic changes of the fluctuation. In operation, first, according to the trend curve obtained by regression analysis, the heart rate change rate at each moment is calculated to obtain the instantaneous rate data of heart rate fluctuation. Then, by accumulating these instantaneous rate data, the trend quantification data of heart rate fluctuation is obtained. This process can also be further evaluated by introducing statistical methods (such as variance analysis, standard deviation calculation, etc.) to evaluate the regularity and consistency of fluctuations. Ultimately, the quantitative data of the disordered scattered trend of heart rate obtained provides an accurate quantitative basis for subsequent fatigue assessment and rehabilitation intervention. These data can reveal the changes in the heart rate of patients under different exercise intensities and fatigue states, and thus provide support for personalized rehabilitation plans.

[0109] Preferably, step S233 includes the following steps:

[0110] Perform cluster feature extraction processing on the multi-dimensional scattered point data of extreme point distribution to obtain extreme point cluster feature data;

[0111] Perform cluster variance calculation on extreme point cluster feature data to obtain extreme point cluster variance data;

[0112] According to the extreme point clustering variance data, disordered scatter point regression analysis is performed on the extreme point distribution multidimensional scatter point data to obtain the heart rate disordered scatter point trend regression data.

[0113] @First, cluster feature extraction is performed on the multi-dimensional scattered data of extreme point distribution. The extreme point distribution data reflects the temporal relationship and characteristics of each extreme point in the heart rate waveform, so it is necessary to extract the common features in the data through cluster analysis. In this process, the multi-dimensional features of each extreme point (such as time position, amplitude, relationship with adjacent extreme points, etc.) are first used as the input parameters of clustering, and these extreme points are grouped using unsupervised learning algorithms (such as K-means clustering algorithm, DBSCAN, etc.). Each cluster represents a typical pattern or fluctuation stage in the heart rate waveform. In the specific operation, the feature vector of each extreme point is first calculated. The feature vector contains multiple aspects of heart rate fluctuation, such as peak position, distance between peaks, heart rate change rate, etc. Then, the clustering algorithm is used to group these feature vectors to obtain multiple categories, and the extreme points in each category have similar characteristics. The core idea of ​​the clustering algorithm is to ensure that the extreme points in each cluster are as similar as possible by minimizing the intra-class distance, while the extreme points between different clusters have significant differences. Finally, the extreme point cluster feature data obtained through this step can provide effective input information for subsequent regression analysis. Based on the extreme point cluster feature data obtained in step S233, cluster variance calculation is performed. The purpose of calculating the cluster variance is to quantify the degree of discreteness of the data points in each cluster, reflecting the internal consistency of each extreme point group. The calculation of the variance can be completed by the following steps: first, the mean of all extreme points in each cluster is calculated (i.e., the center of the cluster), and then the Euclidean distance between each extreme point and the cluster center is calculated, and the squares of these distances are averaged to obtain the variance of the cluster. In specific operations, the calculation of cluster variance can be divided into two steps. The first step is to calculate the average eigenvalue of the extreme points in each cluster for each cluster to obtain the center point of the cluster. The second step is to calculate the distance between each extreme point in the cluster and the cluster center, obtain the difference between each extreme point and the center, and then find the cluster variance. The cluster variance reflects the consistency within each extreme point cluster. The smaller the variance, the more similar the features between the extreme points in the cluster, and the higher the cluster quality. The result of the variance calculation provides quantitative cluster features for subsequent regression analysis, which further helps to judge the importance and role of each cluster in heart rate fluctuations. Using the extreme point cluster variance data obtained in step S234, an unordered scattered point regression analysis is performed on the multidimensional scattered point data of the extreme point distribution. The goal of this process is to establish the relationship between the extreme point features and the heart rate change trend through regression analysis, and further quantify the law of heart rate fluctuations. First, the variance data of each cluster is used as the input parameter of the regression analysis, and the multidimensional features of each extreme point (such as peak time, amplitude, etc.) are combined for regression modeling. The relationship between the extreme point features and the heart rate fluctuation is fitted by regression methods such as the least squares method to obtain a regression model.In the regression process, considering the disorder of scattered data, the regression method should adopt algorithms that adapt to disordered data, such as local weighted regression (LOWESS) or non-parametric regression, to ensure that the regression model can more accurately reflect the trend of scattered data. The result of regression analysis is a set of regression parameters that describe the relationship between extreme point clustering variance data and heart rate fluctuation trend. These regression data provide a quantitative basis for the trend of heart rate fluctuation for subsequent steps, and can further reveal the regularity of heart rate fluctuation and its relationship with the degree of exercise fatigue. Finally, the disordered scattered trend regression data of heart rate obtained through this step will provide data support for the next step of quantitative analysis.

[0114] Preferably, step S24 includes the following steps:

[0115] Step S241: performing trend rising boundary segmentation processing on the disordered scattered point trend quantization data of the heart rate to obtain the heart rate trend rising boundary segmentation data;

[0116] Step S242: performing inflection point distribution space identification on the disordered scattered point trend quantization data of the heart rate according to the heart rate trend rising boundary segmentation data to obtain inflection point distribution space identification data;

[0117] Step S243: Calculate the distribution time series difference of the inflection point distribution space identification data to obtain the inflection point distribution space time series difference;

[0118] Step S244: performing inflection point dynamic distribution fitting on the disordered scattered trend quantification data of the heart rate according to the inflection point distribution spatial time series difference to obtain trend dynamic inflection point distribution fitting data.

[0119] @First, the disordered scattered trend quantification data of heart rate is segmented for the rising boundary of the trend. The purpose of this step is to identify the key points of the trend change in the heart rate data, especially the starting stage of the heart rate rise. In the specific operation, the disordered scattered trend quantification data of heart rate is first smoothed to eliminate some instantaneous fluctuations and noise. Common smoothing methods include median filtering or Gaussian filtering, which can effectively eliminate high-frequency noise and highlight the long-term trend of the data. Then, by setting a certain threshold, the change points from the stable state to the rapid rising state in the heart rate data are detected, and these change points are the rising boundary of the trend. The specific detection method can be based on gradient calculation, that is, calculating the rate of change of the heart rate signal at different time points. When the rate of change exceeds the set threshold, the point is considered to be the rising boundary. After segmentation, the entire heart rate data is divided into two main parts: one is the stable stage before the rise, and the other is the trend stage after the rise. The "heart rate trend rising boundary segmentation data" obtained by this step will provide a basis for the subsequent inflection point analysis and fitting. According to the "heart rate trend rising boundary segmentation data" obtained in step S241, the inflection point distribution space of the disordered scattered trend quantification data of heart rate is identified. In the specific operation, firstly, the data segmented in the previous step is used to define all scattered data in the heart rate rising stage. Then, the second-order difference is performed on the data segment to calculate the acceleration change of each data point. When the acceleration changes significantly, the point is considered to be an inflection point. At this time, a curve fitting method, such as polynomial fitting or spline fitting, can be used to more accurately identify the position of the inflection point. Through this method, not only the specific position of the inflection point can be determined, but also the trend direction of the inflection point can be analyzed, that is, whether the inflection point changes upward or downward. The identification of the inflection point distribution space is based on the temporal distribution law of these inflection points, which further reveals the key change interval in the heart rate change. Finally, the obtained "inflection point distribution space identification data" describes the spatial distribution characteristics of the inflection points of the heart rate signal in the trend rising stage, and provides data support for the next step of the time difference calculation. Based on the "inflection point distribution space identification data" obtained in step S242, the inflection point distribution space time difference is calculated. The purpose of this step is to quantify the time difference between the inflection points, and then analyze the distribution characteristics of the inflection points on the time axis. In the specific operation, first extract all the inflection point timestamps identified in step S242, and then calculate the time difference between these timestamps. A simple time difference calculation formula can be used to calculate the time interval between two adjacent inflection points. Statistical analysis is performed on the time differences of all inflection points to obtain the timing differences between each inflection point and the adjacent inflection point. These timing differences reflect the frequency changes in heart rate fluctuations. Inflection points with shorter time intervals indicate more drastic changes in heart rate, while inflection points with longer time intervals indicate a more stable change trend. The calculated "inflection point distribution space timing difference" data can provide timing feature support for the subsequent inflection point dynamic distribution fitting.Based on the "inflection point distribution spatial time series difference" data obtained in step S243, the inflection point dynamic distribution is fitted to the disordered scattered point trend quantitative data of the heart rate. The purpose of this step is to establish the relationship between the disordered scattered point data of the heart rate and the dynamic distribution of the inflection point through the fitting method, and further analyze the law of heart rate fluctuation. In the specific operation, firstly, the time series difference data obtained in the previous step is used, combined with the position of each inflection point and its adjacent time difference, and the inflection point and its time distribution are fitted by a dynamic fitting method (such as weighted least squares method). During the fitting process, the weight can be adjusted according to the time difference between the inflection points, and the inflection points with shorter time intervals are given higher weights to reflect their importance in the trend change. The goal of fitting is to accurately capture the dynamic characteristics of heart rate fluctuations through a mathematical model, including the trend of heart rate changes, fluctuation amplitude and periodic characteristics.

[0120] Preferably, step S3 comprises the following steps:

[0121] Step S31: performing multi-scale rule intersection analysis on fatigue heart rate change data according to trend dynamic inflection point distribution fitting data to obtain fatigue heart rate change rule intersection data;

[0122] Step S32: performing resting heart rate difference evaluation on the fatigue heart rate change rule intersection data to obtain resting heart rate difference evaluation data;

[0123] Step S33: simulating the heart fatigue level according to the intersection data of fatigue heart rate change rules and the resting heart rate difference evaluation data to obtain heart fatigue level simulation data;

[0124] Step S34: Perform a quantitative assessment of rehabilitation deviation based on the cardiac fatigue level simulation data, the resting heart rate difference assessment data, and the fatigue heart rate change rule intersection data to obtain rehabilitation deviation assessment data.

[0125] As an example of the present invention, refer to Figure 3 As shown, in this example, step S3 includes:

[0126] Step S31: performing multi-scale rule intersection analysis on fatigue heart rate change data according to trend dynamic inflection point distribution fitting data to obtain fatigue heart rate change rule intersection data;

[0127] @According to the "trend dynamic inflection point distribution fitting data" obtained in step S24, a multi-scale law intersection analysis is performed on the fatigue heart rate change data to obtain the fatigue heart rate change law intersection data. In the specific implementation, the trend information of multiple scales is first extracted from the trend dynamic inflection point distribution fitting data, usually by decomposing the heart rate change signal, and the signal can be decomposed at multiple scales using methods such as wavelet transform (such as discrete wavelet transform, DWT). Through this method, the heart rate signal is separated from different frequency bands to obtain the heart rate fluctuation law at different time scales. The signal of each scale reflects the changes of different frequency components, and the low-frequency component usually corresponds to a slower trend change, while the high-frequency component corresponds to a fast fluctuation. Next, using intersection analysis, these different scales of fluctuation information are cross-analyzed to identify time periods that show similar laws at multiple scales. These common parts are the main laws of fatigue heart rate changes. The "fatigue heart rate change law intersection data" obtained in this step reflects the multi-scale law of heart rate changes under fatigue state, and provides a comprehensive basis for the characteristics of heart rate changes in subsequent steps.

[0128] Step S32: performing resting heart rate difference evaluation on the fatigue heart rate change rule intersection data to obtain resting heart rate difference evaluation data;

[0129] @ Perform resting heart rate difference evaluation on the "intersection data of fatigue heart rate change rules" obtained in step S31 to obtain resting heart rate difference evaluation data. In the specific implementation, firstly, the intersection data of fatigue heart rate change rules is analyzed in time domain and frequency domain. Resting heart rate is usually the heart rate baseline data measured in a completely resting state. Therefore, it is first necessary to extract the heart rate signal in the resting state from the fatigue heart rate data. In order to accurately extract the resting heart rate data, the resting interval recognition method of the electrocardiogram signal can be used, such as the recognition of the continuous stable period based on the electrocardiogram signal, by setting a stable threshold or using the sliding window technology to identify the period with small heart rate change. Through this method, after obtaining the resting heart rate data, it is compared with the heart rate data in the fatigue state to calculate the difference between the two. The difference evaluation is not just a simple numerical comparison, but through time series difference analysis, peak ratio analysis and other methods, the differences between the two in terms of fluctuation amplitude, frequency change, etc. are deeply analyzed. Finally, the obtained "resting heart rate difference evaluation data" reflects the amplitude of heart rate change under different states, which provides an important quantitative indicator for further fatigue evaluation and rehabilitation analysis.

[0130] Step S33: simulating the heart fatigue level according to the intersection data of fatigue heart rate change rules and the resting heart rate difference evaluation data to obtain heart fatigue level simulation data;

[0131] @According to the "intersection data of fatigue heart rate change rules" obtained in step S31 and the "resting heart rate difference evaluation data" obtained in step S32, perform heart fatigue level simulation to obtain heart fatigue level simulation data. In the specific implementation, first use the intersection data of fatigue heart rate change rules and the resting heart rate difference evaluation data to construct a multidimensional feature space. Each dimension of this space represents a different physiological parameter, such as heart rate fluctuation, peak change, frequency characteristics, etc. After constructing the feature space, these data are modeled using a nonlinear function fitting method to predict the heart fatigue level. During the fitting process, methods such as exponential regression, linear regression or polynomial regression can be used to construct a regression model based on different features. The heart fatigue level obtained by simulation can quantitatively evaluate the individual's fatigue level based on the law of heart rate change. Finally, through fatigue level simulation, "heart fatigue level simulation data" is obtained, which reflects the individual's fatigue level under different exercise and resting states, and serves as an important basis for heart health assessment.

[0132] Step S34: performing a quantitative assessment of rehabilitation deviation based on the cardiac fatigue level simulation data, the resting heart rate difference assessment data, and the fatigue heart rate change rule intersection data to obtain rehabilitation deviation assessment data;

[0133] @Based on the "heart fatigue level simulation data" obtained in step S33, the "resting heart rate difference evaluation data" obtained in step S32, and the "fatigue heart rate change law intersection data" obtained in step S31, a quantitative evaluation of rehabilitation deviation is performed to obtain rehabilitation deviation evaluation data. In the specific implementation, each data set is first analyzed jointly. The goal of rehabilitation deviation evaluation is to quantify the difference between an individual in the cardiac rehabilitation process and the predetermined health state. In the analysis process, the "heart fatigue level simulation data", "resting heart rate difference evaluation data" and "fatigue heart rate change law intersection data" are combined by multivariate statistical analysis methods (such as principal component analysis PCA or canonical correlation analysis CCA) to extract the key factors affecting the rehabilitation process. These factors include indicators such as heart rate recovery time and heart rate change rate. Combined with these indicators, a rehabilitation deviation model based on time series is established, and the deviation value between the current state and the health goal is calculated by analyzing the change trend during the rehabilitation process. This deviation value reflects the progress and problems in the cardiac rehabilitation process. Finally, the obtained "rehabilitation deviation evaluation data" will provide a scientific basis for the subsequent optimization of rehabilitation strategies and help formulate more accurate rehabilitation plans.

[0134] Preferably, step S33 includes the following steps:

[0135] Step S331: extracting the intersection feature vector of the heart rate change law from the intersection data of the fatigue heart rate change law to obtain the intersection vector of the fatigue heart rate law;

[0136] Step S332: performing multimodal fusion processing on the resting heart rate difference evaluation data according to the fatigue heart rate rule intersection vector to obtain fatigue heart rate multimodal fusion data;

[0137] Step S333: Perform heart fatigue level simulation based on fatigue heart rate multimodal fusion data to obtain heart fatigue level simulation data.

[0138] @ Extract the intersection feature vector of the heart rate change law from the "intersection data of fatigue heart rate change law" obtained in step S31 to obtain the "intersection vector of fatigue heart rate law". First, by analyzing the time series characteristics of the intersection data of fatigue heart rate change law, multiple characteristic indicators related to heart rate change are extracted. This process involves the time series characteristics of heart rate, including but not limited to the change amplitude, frequency distribution, peak fluctuation, acceleration and deceleration rate, etc. The signal can be decomposed in the frequency domain by Fourier transform (FFT) or wavelet transform (DWT) to identify the heart rate change pattern in different frequency bands, especially the specific change trend shown in the fatigue state. Through these frequency domain features, a set of representative features can be obtained to form a feature vector, which represents the main characteristics of the heart rate change law. In the extraction process, feature selection technology (such as variance analysis or information gain) is further used to remove redundant features, so as to obtain a concise and efficient feature vector. This "intersection vector of fatigue heart rate law" can fully reflect the change pattern of heart rate in fatigue state and provide a concise expression for subsequent steps. According to the "fatigue heart rate law intersection vector" obtained in step S331 and the "resting heart rate difference evaluation data" obtained in step S32, multimodal fusion processing is performed to obtain "fatigue heart rate multimodal fusion data". First of all, the goal of multimodal data fusion is defined as the time domain characteristics of the comprehensive heart rate change law and the physiological characteristics of the resting heart rate difference. The two represent different physiological signal dimensions, so it is necessary to integrate these data through appropriate fusion algorithms to achieve information complementarity. In the specific operation process, the "resting heart rate difference evaluation data" is first standardized so that it is in the same dimensional range as the "fatigue heart rate law intersection vector". Then, the key features of the two types of data are fused together using the weighted summation method or the mutual information fusion method. The weighted summation method reflects the relative importance of each type of data by setting weights for each type of data. The weights can be adjusted by variance analysis or cross-validation, so that data from different sources have a reasonable impact on the final result. After multimodal fusion processing, the obtained "fatigue heart rate multimodal fusion data" contains the heart rate change information obtained from different angles, reflects the comprehensive physiological characteristics under fatigue state, and provides complete data support for further simulation of cardiac fatigue level. Based on the "fatigue heart rate multimodal fusion data" obtained in step S332, cardiac fatigue level simulation is performed to obtain "cardiac fatigue level simulation data". First, according to the fusion data obtained in the previous step, a fatigue level model based on regression analysis or statistical inference is established to predict the degree of cardiac fatigue. In the process of model construction, it is necessary to determine the mathematical relationship between fatigue level and characteristics such as heart rate fluctuations, resting heart rate differences, and fatigue heart rate changes through historical data or clinical research results.In order to enhance the accuracy of the model, weighted regression models, principal component analysis (PCA) and other methods can be used to process the fused data, reduce the multidimensional data to the main components, and then perform regression analysis to estimate the level of cardiac fatigue. Specifically, the regression model can calculate a value representing the degree of cardiac fatigue through linear or nonlinear regression methods, combined with the amplitude of fatigue heart rate fluctuations, resting heart rate differences and other physiological characteristics. After model fitting, the obtained "heart fatigue level simulation data" accurately reflects the physiological load of the heart under fatigue, and can be used to monitor the patient's fatigue condition and adjust the rehabilitation plan. This data provides a key quantitative basis for subsequent rehabilitation assessment and cardiac rehabilitation intervention.

[0139] Preferably, step S34 includes the following steps:

[0140] Step S341: performing differential feature extraction on the resting heart rate difference assessment data according to the heart fatigue level simulation data to obtain heart rate differential feature data;

[0141] Step S342: performing polynomial fitting analysis on the heart rate difference feature data and the intersection data of fatigue heart rate variation law to obtain the fatigue heart rate polynomial fitting coefficient;

[0142] Step S343: performing nonlinear mapping transformation on the heart rate difference characteristic data according to the fatigue heart rate polynomial fitting coefficient to obtain the heart rate nonlinear characteristic mapping data;

[0143] Step S344: Perform quantitative assessment of rehabilitation deviation based on the heart rate nonlinear characteristic mapping data, fatigue heart rate polynomial fitting coefficients and heart rate differential characteristic data to obtain rehabilitation deviation assessment data.

[0144] @According to the "heart fatigue level simulation data" obtained in step S333, the "resting heart rate difference evaluation data" obtained in step S32 is subjected to differential feature extraction to obtain "heart rate differential feature data". Specifically, the resting heart rate difference data is first subjected to time series analysis to capture its changing trend. The purpose of differential feature extraction is to highlight the instantaneous rate of change of the heart rate by eliminating the basic fluctuations of the resting heart rate. During the operation, the resting heart rate difference data is first subjected to one or more differential processing to obtain the rate information of the heart rate change. The first order difference can be obtained by calculating the difference between adjacent data points, or a higher order difference can be calculated to capture more subtle changes. In addition, local differential processing is performed on different time windows to further identify the characteristics of short-term fluctuations. The "heart rate differential feature data" obtained through this process can effectively reveal the instantaneous fluctuation characteristics of the resting heart rate, and thus provide basic information on the dynamic changes of the heart rate for subsequent steps. Based on the "heart rate differential feature data" obtained in step S341 and the "fatigue heart rate change law intersection data" obtained in step S31, a polynomial fitting analysis is performed to obtain the "fatigue heart rate polynomial fitting coefficient". In the specific operation, the relationship between the heart rate differential feature data and the fatigue heart rate change law intersection data is first modeled. The polynomial fitting method is used to capture the nonlinear relationship between the heart rate differential feature and the fatigue heart rate change law. Select a suitable polynomial order, such as a quadratic, cubic or quartic polynomial, to fit the data. Solve the polynomial coefficients by the least squares method or the weighted least squares method to obtain a best fitting polynomial expression. In the fitting process, cross-validation or validation sets are used to avoid overfitting and ensure the robustness of the model. The "fatigue heart rate polynomial fitting coefficient" obtained by polynomial fitting reflects the nonlinear dependency between the heart rate differential feature and the fatigue heart rate change law, and can provide necessary parameters for subsequent nonlinear mapping transformations. According to the "fatigue heart rate polynomial fitting coefficient" obtained in step S342, the "heart rate differential feature data" obtained in step S341 is subjected to a nonlinear mapping transformation to obtain "heart rate nonlinear feature mapping data". First, the heart rate differential feature data is mapped through a polynomial function using the polynomial fitting coefficient obtained in step S342. The purpose of this process is to convert the linearly changing heart rate differential data into nonlinear features through a polynomial relationship, so that more complex heart rate change patterns can be revealed. Specifically, for each differential feature data, the obtained polynomial model is substituted into the calculation to obtain a new nonlinear value. In order to enhance the accuracy of the nonlinear mapping, the nonlinear mapping effect can be further optimized by adjusting the fitting coefficient or using a polynomial function with a regularization term.In addition, the local weighting method can be used in the mapping process to dynamically adjust according to the change trend of the data points to ensure that the nonlinear mapping can adapt to different data patterns. Finally, the obtained "heart rate nonlinear feature mapping data" has a high expressive ability, can more accurately describe the nonlinear characteristics in the heart rate change, and provide more refined data input for the subsequent quantitative evaluation of rehabilitation deviation. Based on the "heart rate nonlinear feature mapping data" obtained in step S343, the "fatigue heart rate polynomial fitting coefficient" obtained in step S342, and the "heart rate differential feature data" obtained in step S341, a quantitative evaluation of rehabilitation deviation is performed to obtain "rehabilitation deviation evaluation data". First, the quantitative indicators of rehabilitation deviation evaluation are set, which usually involve multiple aspects such as heart rate recovery, heart rate fluctuation amplitude, and resting heart rate recovery level. The feature data obtained from the nonlinear mapping is compared with the heart rate data in the normal rehabilitation process to analyze the deviation between the patient's current heart rate state and the ideal heart rate state. Through the polynomial fitting coefficient and the heart rate differential feature data, a deviation calculation model is constructed, which compares the patient's heart rate nonlinear characteristics with the normal healthy heart rate data to calculate the deviation value. The specific deviation calculation method can quantitatively evaluate the degree of heart rate deviation through weighted difference, square error, absolute error and other methods. In addition, dynamic thresholds can be set to adjust the deviation according to individual differences of patients, making the evaluation results more personalized. On this basis, through further analysis of the evaluation data, specific "rehabilitation deviation evaluation data" are obtained. This data provides the necessary quantitative basis for the adjustment of the rehabilitation process and provides data support for the formulation of personalized rehabilitation plans.

[0145] Preferably, the present invention further provides an ECG monitoring system based on cardiac rehabilitation data, which is used to perform the ECG monitoring method based on cardiac rehabilitation data as described above. The ECG monitoring system based on cardiac rehabilitation data comprises:

[0146] The ECG signal mapping module is used to collect ECG signals of patients in different motion states through a multi-lead ECG instrument and a motion monitoring device to obtain ECG signals in different motion states; perform motion state fatigue ECG signal mapping on the ECG signals in different motion states to obtain fatigue ECG mapping signals;

[0147] The inflection point dynamic distribution fitting module is used to analyze the heart rate change of the fatigue ECG mapping signal to obtain fatigue heart rate change data; to quantify the disordered scattered point trend of the fatigue heart rate change data to obtain the disordered scattered point trend quantification data of the heart rate; to perform inflection point dynamic distribution fitting on the disordered scattered point trend quantification data of the heart rate to obtain the trend dynamic inflection point distribution fitting data;

[0148] The rehabilitation deviation quantitative assessment module is used to perform multi-scale rule intersection analysis on fatigue heart rate change data according to the trend dynamic inflection point distribution fitting data to obtain fatigue heart rate change rule intersection data; perform rehabilitation deviation quantitative assessment based on fatigue heart rate change rule intersection data to obtain rehabilitation deviation assessment data;

[0149] The terminal feedback module is used to send the rehabilitation deviation assessment data to the terminal to perform ECG monitoring of cardiac rehabilitation data.

[0150] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0151] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. An electrocardiogram monitoring method based on cardiac rehabilitation data, characterized in that: The following steps are involved: Step S1: collecting ECG signals of the patient in different motion states by means of a multi-lead electrocardiograph and a motion monitoring device to obtain ECG signals in different motion states; performing motion state fatigue ECG signal mapping on the ECG signals in different motion states to obtain fatigue ECG mapping signals; Step S2: analyzing the heart rate variation of the fatigue ECG mapping signal to obtain fatigue heart rate variation data; performing disordered scattered point trend quantification on the fatigue heart rate variation data to obtain disordered scattered point trend quantification data of the heart rate; Perform inflection point dynamic distribution fitting on the disordered scattered trend quantitative data of heart rate to obtain trend dynamic inflection point distribution fitting data; Step S3: performing multi-scale rule intersection analysis on fatigue heart rate change data according to trend dynamic inflection point distribution fitting data to obtain fatigue heart rate change rule intersection data; Based on the intersection data of fatigue heart rate change rules, quantitative assessment of rehabilitation deviation is performed to obtain rehabilitation deviation assessment data. Step S3 includes the following steps: Step S31: performing multi-scale rule intersection analysis on fatigue heart rate change data according to trend dynamic inflection point distribution fitting data to obtain fatigue heart rate change rule intersection data; Step S32: performing resting heart rate difference evaluation on the fatigue heart rate change rule intersection data to obtain resting heart rate difference evaluation data; Step S33: simulating the heart fatigue level according to the intersection data of fatigue heart rate change rules and the resting heart rate difference evaluation data to obtain heart fatigue level simulation data; Step S34: quantitatively evaluate the rehabilitation deviation based on the heart fatigue level simulation data, the resting heart rate difference evaluation data, and the fatigue heart rate change rule intersection data to obtain rehabilitation deviation evaluation data. Step S34 includes the following steps: Step S341: performing differential feature extraction on the resting heart rate difference assessment data according to the heart fatigue level simulation data to obtain heart rate differential feature data; Step S342: performing polynomial fitting analysis on the heart rate difference feature data and the intersection data of fatigue heart rate variation law to obtain the fatigue heart rate polynomial fitting coefficient; Step S343: performing nonlinear mapping transformation on the heart rate difference characteristic data according to the fatigue heart rate polynomial fitting coefficient to obtain the heart rate nonlinear characteristic mapping data; Step S344: performing quantitative assessment of rehabilitation deviation based on the heart rate nonlinear characteristic mapping data, fatigue heart rate polynomial fitting coefficients and heart rate differential characteristic data to obtain rehabilitation deviation assessment data; Step S4: Send the rehabilitation deviation assessment data to the terminal to perform ECG monitoring of cardiac rehabilitation data.

2. The ECG monitoring method based on cardiac rehabilitation data according to claim 1, characterized in that: Step S1 includes the following steps: Step S11: collecting ECG signals of the patient in different motion states by using a multi-lead electrocardiograph and a motion monitoring device to obtain ECG signals in different motion states; Step S12: performing noise suppression processing on the ECG signals under different motion states to obtain ECG denoised signals under different motion states; Step S13: performing time-series segmentation processing on the ECG denoising signals under different motion states to obtain ECG denoising time-series segmented signals; Step S14: performing motion state fatigue ECG signal mapping on the ECG denoising signals under different motion states according to the ECG denoising time sequence segmented signals to obtain fatigue ECG mapping signals.

3. The ECG monitoring method based on cardiac rehabilitation data according to claim 1, characterized in that: Step S2 includes the following steps: Step S21: analyzing the heart rate variation of the fatigue ECG mapping signal to obtain fatigue heart rate variation data; Step S22: performing waveform time domain decomposition on the fatigue heart rate variation data to obtain fatigue heart rate waveform time domain data; Step S23: performing disordered scattered point trend quantification on fatigue heart rate change data according to fatigue heart rate waveform time domain data to obtain disordered scattered point trend quantification data of heart rate; Step S24: performing inflection point dynamic distribution fitting on the disordered scattered trend quantified data of the heart rate to obtain trend dynamic inflection point distribution fitting data.

4. The ECG monitoring method based on cardiac rehabilitation data according to claim 3, characterized in that: Step S23 includes the following steps: Step S231: performing region segmentation processing on the fatigue heart rate waveform time domain data to obtain heart rate time domain region segmentation data; Step S232: performing extreme point distribution multi-dimensional scattered point matching on the fatigue heart rate waveform time domain data according to the heart rate time domain region segmentation data to obtain extreme point distribution multi-dimensional scattered point data; Step S233: performing disordered scatter point regression analysis on the multi-dimensional scatter point data of extreme point distribution to obtain disordered scatter point trend regression data of heart rate; Step S234: performing disordered scatter trend quantification on fatigue heart rate variation data according to the disordered scatter trend regression data of heart rate to obtain disordered scatter trend quantification data of heart rate.

5. The ECG monitoring method based on cardiac rehabilitation data according to claim 4, characterized in that: Step S233 includes the following steps: Perform cluster feature extraction processing on the multi-dimensional scattered point data of extreme point distribution to obtain extreme point cluster feature data; Perform cluster variance calculation on extreme point cluster feature data to obtain extreme point cluster variance data; According to the extreme point clustering variance data, disordered scatter point regression analysis is performed on the extreme point distribution multidimensional scatter point data to obtain the heart rate disordered scatter point trend regression data.

6. The ECG monitoring method based on cardiac rehabilitation data according to claim 3, characterized in that: Step S24 includes the following steps: Step S241: performing trend rising boundary segmentation processing on the disordered scattered point trend quantization data of the heart rate to obtain the heart rate trend rising boundary segmentation data; Step S242: performing inflection point distribution space identification on the disordered scattered point trend quantization data of the heart rate according to the heart rate trend rising boundary segmentation data to obtain inflection point distribution space identification data; Step S243: Calculate the distribution time series difference of the inflection point distribution space identification data to obtain the inflection point distribution space time series difference; Step S244: performing inflection point dynamic distribution fitting on the disordered scattered trend quantification data of the heart rate according to the inflection point distribution spatial time series difference to obtain trend dynamic inflection point distribution fitting data.

7. The electrocardiogram monitoring method based on cardiac rehabilitation data according to claim 1, characterized in that: Step S33 includes the following steps: Step S331: extracting the intersection feature vector of the heart rate change law from the intersection data of the fatigue heart rate change law to obtain the intersection vector of the fatigue heart rate law; Step S332: performing multimodal fusion processing on the resting heart rate difference evaluation data according to the fatigue heart rate rule intersection vector to obtain fatigue heart rate multimodal fusion data; Step S333: Perform heart fatigue level simulation based on fatigue heart rate multimodal fusion data to obtain heart fatigue level simulation data.

8. An electrocardiogram monitoring system based on cardiac rehabilitation data, characterized in that: Used to perform the ECG monitoring method based on cardiac rehabilitation data as claimed in claim 1, the ECG monitoring system based on cardiac rehabilitation data comprises: The ECG signal mapping module is used to collect ECG signals of patients in different motion states through a multi-lead ECG instrument and a motion monitoring device to obtain ECG signals in different motion states; perform motion state fatigue ECG signal mapping on the ECG signals in different motion states to obtain fatigue ECG mapping signals; The inflection point dynamic distribution fitting module is used to analyze the heart rate change of the fatigue ECG mapping signal to obtain fatigue heart rate change data; to quantify the disordered scattered point trend of the fatigue heart rate change data to obtain the disordered scattered point trend quantification data of the heart rate; to perform inflection point dynamic distribution fitting on the disordered scattered point trend quantification data of the heart rate to obtain the trend dynamic inflection point distribution fitting data; The rehabilitation deviation quantitative assessment module is used to perform multi-scale rule intersection analysis on fatigue heart rate change data according to the trend dynamic inflection point distribution fitting data to obtain fatigue heart rate change rule intersection data; perform rehabilitation deviation quantitative assessment based on fatigue heart rate change rule intersection data to obtain rehabilitation deviation assessment data; The terminal feedback module is used to send the rehabilitation deviation assessment data to the terminal to perform ECG monitoring of cardiac rehabilitation data.

Images