Intelligent heart rate monitoring method and system based on heart sound and electrocardiosignal

Abstract

The invention relates to the technical field of biomedical signal processing and intelligent health monitoring, and particularly discloses an intelligent heart rate monitoring method and system based on heart sound and electrocardiosignals, heart sound signals and electrocardiosignals of a user are collected in real time, quality evaluation is conducted on the collected signals through signal quality indexes, and accurate signals are extracted; extracting time domain and frequency domain features of the first heart sound and the second heart sound of the heart sound signal, fusing multiple features by adopting wavelet transform and a main frequency analysis algorithm, and calculating a heart sound abnormal coefficient to evaluate the abnormal degree of the heart sound signal; the vibration amplitude and frequency characteristics of P waves in the electrocardiosignals are extracted, and an electrocardiograph abnormal coefficient is calculated through a vibration amplitude abnormal coefficient and a vibration frequency fluctuation coefficient and used for evaluating the stability of the electrocardiosignals; based on the heart sound abnormal coefficient and the electrocardiogram abnormal coefficient, a decision tree model is constructed, a heart rate abnormal index is dynamically calculated, dynamic monitoring of the heart rate is achieved, a dynamic early warning mechanism is conducted, and heart rate abnormal information is output.

Description

Technical Field

[0001] The present invention relates to the technical fields of biomedical signal processing and intelligent health monitoring, and particularly relates to an intelligent heart rate monitoring method and system based on heart sound and electrocardiogram signals. Background Art

[0002] With the rapid development of intelligent health monitoring technology, biomedical signal analysis methods based on electrocardiogram signals and heart sound signals have been widely used in cardiac health management. Electrocardiogram signals can reflect the electrophysiological activity characteristics of the heart, while heart sound signals can capture information on the mechanical activity of the heart. The comprehensive analysis of the two provides an important basis for heart rate monitoring and early warning of heart diseases. However, in a complex environment, heart sound signals and electrocardiogram signals are easily affected by noise, interference, and abnormal data, resulting in a decline in signal quality, which has an adverse impact on the accuracy and stability of subsequent analysis. In addition, there are significant differences in signal characteristics among different individuals, and traditional single-feature analysis methods are difficult to comprehensively evaluate the cardiac health status.

[0003] The existing technologies have the following deficiencies:

[0004] In the existing technologies, the processing of heart sound signals and electrocardiogram signals mostly adopts an independent analysis method, lacking a comprehensive analysis means for the characteristics of the two, and it is difficult to accurately reflect the correlation and difference between the signals. At the same time, the existing methods have limited ability to eliminate noise and abnormal signals in signal quality assessment, resulting in abnormal signals being misidentified as normal signals, reducing the reliability of the analysis. In addition, traditional heart rate monitoring methods mostly rely on fixed algorithm models and are difficult to dynamically adapt to the signal characteristics that change in real time, resulting in insufficient accuracy and real-time performance in detecting abnormal heart rates. The above deficiencies make the assessment of cardiac health status at risk of misjudgment and difficult to meet the high-precision and personalized health monitoring requirements. Summary of the Invention

[0005] The purpose of the present invention is to provide an intelligent heart rate monitoring method and system based on heart sound and electrocardiogram signals to solve the problems in the above background.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] An intelligent heart rate monitoring method based on heart sound and electrocardiogram signals, comprising:

[0008] Real-time collecting the heart sound signal and electrocardiogram signal of a user, performing quality assessment on the collected heart sound signal and electrocardiogram signal, determining the accuracy of the data through a signal quality index, eliminating noise interference and abnormal signals, and extracting accurate heart sound signals and electrocardiogram signals;

[0009] Extract the time-domain and frequency-domain features of the first heart sound and the second heart sound in the heart sound signal, perform fusion processing on the time-domain and frequency-domain features, calculate the heart sound abnormality coefficient, and use it to evaluate the abnormality degree of the heart sound signal;

[0010] Extract the vibration features of the P wave in the electrocardiogram signal, including: vibration amplitude feature and vibration frequency feature, process the vibration features, and calculate the electrocardiogram abnormality coefficient through comprehensive processing, which is used to evaluate the stability of the electrocardiogram signal;

[0011] Based on the heart sound abnormality coefficient and the electrocardiogram abnormality coefficient, dynamically calculate the heart rate abnormality index. According to the heart rate abnormality index, divide the user's heart rate into abnormal heart rate and normal heart rate, and trigger a dynamic warning mechanism based on the abnormal heart rate to output heart rate abnormality information.

[0012] Preferably: The process of obtaining the signal quality index is as follows:

[0013] Obtain signal data through a heart sound sensor and an electrocardiogram electrode, including: heart sound signal and electrocardiogram signal;

[0014] Preprocess the collected signal data;

[0015] Obtain the preprocessed signal data;

[0016] Perform further filtering optimization on the heart sound signal and the electrocardiogram signal, dynamically adjust the filter coefficients, obtain the optimized signal data, and calculate the error signal of the signal;

[0017] Calculate the signal-to-noise ratio of the signal data, and the calculation expression is:

[0018]

[0019] In the formula, x(t) represents the preprocessed signal data, n(t) represents the noise signal, represents the expected value, and t represents time;

[0020] Perform a fast Fourier transform on the optimized signal and calculate the spectral energy. The calculation expression is:

[0021]

[0022] In the formula, P(f) represents the spectral energy, f represents the frequency, i represents the i-th frequency, N represents the maximum number of frequencies, and X(f i ) represents the Fourier transform result at the i-th frequency;

[0023] Calculate the ratio of the signal-to-noise ratio of the signal data to the spectral energy to obtain the signal quality index. The calculation expression is:

[0024]

[0025] In the formula, Q Z represents the signal quality index of the signal data.

[0026] Preferably: The extraction of accurate heart sound signals and electrocardiogram signals specifically includes:

[0027] Judging whether the signal quality index is greater than or equal to a preset threshold. If so, it is recorded as an accurate heart sound signal and electrocardiogram signal. If not, it is recorded as an inaccurate heart sound signal and electrocardiogram signal.

[0028] Preferably: The process of obtaining the heart sound abnormality coefficient is as follows:

[0029] Perform denoising processing on the collected heart sound signal. Use the wavelet transform denoising algorithm and the threshold denoising method to remove high-frequency noise and retain the effective heart sound components to obtain the denoised heart sound signal;

[0030] Perform wavelet transform on the denoised heart sound signal to obtain the decomposed wavelet coefficients, and extract the time-domain characteristics of the first heart sound and the second heart sound from the decomposed wavelet coefficients, including the duration, which is calculated through the local extreme values of the wavelet coefficients;

[0031] Among them, the calculation expression of the wavelet transform is:

[0032]

[0033] In the formula, ψ j (t) represents the j-th layer wavelet basis, x1(t) represents the heart sound signal, j represents the number of decomposition layers, J represents the total number of decomposition layers, c j represents the j-th layer wavelet coefficient, and t represents time;

[0034] In the frequency-domain analysis of the wavelet coefficients, extract the frequency-domain characteristics of the heart sound signal, including: the main frequency and the frequency band width; among them, through the spectrum analysis of the wavelet coefficients, calculate the main frequencies of the first heart sound and the second heart sound, and the corresponding frequency band widths;

[0035] Among them, the calculation expression of the main frequency is:

[0036] f peak = argmax f |X(f)|;

[0037] In the formula, f peak represents the main frequency of the heart sound signal, and X(f) is the spectrum of the heart sound signal;

[0038] The frequency band width is obtained by calculating the energy distribution in different frequency bands;

[0039] Fuse the extracted time-domain features and frequency-domain features to calculate a comprehensive feature vector. The calculation expression is as follows:

[0040] F fusion = α·F time + β·F freq ;

[0041] In the formula, F fusion represents the comprehensive feature vector, F time represents the time-domain feature vector, F freq represents the frequency-domain feature vector, and α and β are preset proportional coefficients;

[0042] Calculate the heart sound abnormality coefficient according to the fused features. The calculation expression is as follows:

[0043]

[0044] In the formula, EC represents the heart sound abnormality coefficient, F normal represents the feature vector of normal heart sound, and σ normal represents the standard deviation of the normal heart sound feature.

[0045] Preferably: The processing of the vibration features specifically includes:

[0046] Extract the vibration amplitude feature and vibration frequency feature of the P wave in the electrocardiogram signal. Calculate the vibration amplitude abnormality coefficient according to the vibration amplitude feature, and calculate the vibration frequency fluctuation coefficient according to the fluctuation degree of the vibration frequency feature;

[0047] Among them, the acquisition process of the vibration amplitude abnormality coefficient is as follows:

[0048] Extract the P wave signal from the electrocardiogram signal;

[0049] Perform vibration amplitude calculation on the extracted P wave signal to obtain the vibration amplitude. The calculation expression is as follows:

[0050] A P = max(x Pwave (t)) - min(x Pwave (t));

[0051] In the formula, A P represents the vibration amplitude, x Pwave (t) represents the P wave signal, and t represents time;

[0052] Based on a group of normal P wave signals, calculate the mean and standard deviation of the normal vibration amplitude, and then calculate the vibration amplitude abnormality coefficient of the current P wave. The calculation expression is as follows:

[0053]

[0054] In the formula, ECA Denotes the vibration amplitude anomaly coefficient, μ normal Denotes the mean value of the vibration amplitude, σ normal Denotes the standard deviation of the vibration amplitude;

[0055] Among them, the acquisition process of the vibration frequency fluctuation coefficient is as follows:

[0056] Extract the P-wave signal from the electrocardiogram signal;

[0057] Use the fast Fourier transform to perform frequency-domain analysis on the P-wave signal and extract the main frequency of the P-wave;

[0058] Calculate the fluctuation amplitude of the P-wave frequency, and the calculation expression is:

[0059] Δf Pwave =max(f Pwave (t))-min(f Pwave (t));

[0060] In the formula, Δf Pwave Denotes the fluctuation amplitude of the P-wave frequency, f Pwave (t) denotes the P-wave frequency extracted at different time points;

[0061] Based on the standard deviation of the P-wave frequency, calculate the vibration frequency fluctuation coefficient, and the calculation expression is:

[0062]

[0063] In the formula, EC f Denotes the vibration frequency fluctuation coefficient, σ f Denotes the standard deviation of the P-wave frequency.

[0064] Preferably: the acquisition process of the electrocardiogram anomaly coefficient is as follows:

[0065] Obtain the vibration amplitude anomaly coefficient and the vibration frequency fluctuation coefficient of the P-wave in the electrocardiogram signal, and perform normalization processing on the vibration amplitude anomaly coefficient and the vibration frequency fluctuation coefficient to calculate the electrocardiogram anomaly coefficient.

[0066] Preferably: based on the heart sound anomaly coefficient and the electrocardiogram anomaly coefficient, dynamically calculate the heart rate anomaly index, specifically including:

[0067] Real-time collect the heart sound signal and electrocardiogram signal of the user, and calculate and obtain the heart sound anomaly coefficient and the electrocardiogram anomaly coefficient respectively;

[0068] Construct a decision tree model based on the historical data set, and the historical data set includes the heart sound anomaly coefficient, the electrocardiogram anomaly coefficient and the corresponding heart rate anomaly index label;

[0069] The decision tree selects the heart sound anomaly coefficient and the electrocardiogram anomaly coefficient as features, selects split points based on splitting criteria, and recursively constructs a decision tree model;

[0070] For the newly collected real-time heart sound anomaly coefficient and electrocardiogram anomaly coefficient, recursive decision-making is performed according to the decision tree model through layer-by-layer judgment rules;

[0071] The heart rate anomaly index is output according to the recursive decision-making, and the heart rate anomaly index is a continuous value;

[0072] By dynamically inputting new heart sound and electrocardiogram signal data, the decision tree model is updated in real time, further improving the accuracy and real-time performance of the calculation results.

[0073] Preferably: According to the heart rate anomaly index, the user's heart rate is divided into abnormal heart rate and normal heart rate, specifically including:

[0074] Calculate the mean value of the heart rate anomaly index within a period of time monitored by the user, and determine whether the heart rate anomaly index within the current time period is greater than or equal to a preset threshold. If so, the user's heart rate within the corresponding time period is abnormal and is recorded as an abnormal heart rate. If not, the user's heart rate within the corresponding time period is normal and is recorded as a normal heart rate.

[0075] Furthermore, an intelligent heart rate monitoring system based on heart sound and electrocardiogram signals is proposed, which is used to implement the intelligent heart rate monitoring method based on heart sound and electrocardiogram signals as described above, including:

[0076] A data acquisition module, which acquires the user's heart sound signal and electrocardiogram signal in real time;

[0077] A signal quality evaluation module, which evaluates the quality of the acquired heart sound signal and electrocardiogram signal, determines the accuracy of the data through a signal quality index, eliminates noise interference and abnormal signals, and extracts accurate heart sound signals and electrocardiogram signals;

[0078] A heart sound signal evaluation module, which extracts the time-domain features and frequency-domain features of the first heart sound and the second heart sound in the heart sound signal, performs fusion processing on the time-domain features and frequency-domain features, and calculates a heart sound anomaly coefficient to evaluate the abnormal degree of the heart sound signal;

[0079] An electrocardiogram signal evaluation module, which extracts the vibration features of the P wave in the electrocardiogram signal, including: vibration amplitude features and vibration frequency features, processes the vibration features, and calculates an electrocardiogram anomaly coefficient through comprehensive processing to evaluate the stability of the electrocardiogram signal;

[0080] Heart rate abnormality dynamic judgment module. The heart rate abnormality dynamic judgment module dynamically calculates the heart rate abnormality index based on the heart sound abnormality coefficient and the electrocardiogram abnormality coefficient, divides the user's heart rate into abnormal heart rate and normal heart rate according to the heart rate abnormality index, and triggers a dynamic warning mechanism based on the abnormal heart rate to output heart rate abnormality information.

[0081] Advantages of the present invention:

[0082] (1) By introducing a comprehensive evaluation method of signal quality index, the present invention systematically processes the collected heart sound signals and electrocardiogram signals, eliminates environmental noise and abnormal signal interference, and ensures the purity and accuracy of signal data. Further, through the fusion analysis of the time-domain and frequency-domain characteristics of heart sound signals, combined with wavelet transform denoising algorithm, main frequency extraction and frequency band width evaluation, the capture accuracy of heart sound abnormality characteristics is comprehensively improved, providing multi-dimensional support for abnormal detection. Aiming at the P-wave characteristics in electrocardiogram signals, the present invention innovatively proposes a two-parameter analysis method based on the vibration amplitude abnormality coefficient and the vibration frequency fluctuation coefficient, comprehensively evaluating the stability and dynamic change trend of the signals. Overall, the signal processing method of the present invention shows excellent robustness and adaptability in a complex and changeable signal environment, not only improving the accuracy of signal analysis, but also providing scientific and efficient technical support for heart rate abnormality detection, further laying a reliable data foundation and supporting the intelligent and real-time warning capabilities of heart rate monitoring.

[0083] (2) Through the comprehensive analysis of the heart sound abnormality coefficient and the electrocardiogram abnormality coefficient, the present invention innovatively constructs a dynamic calculation method of heart rate abnormality index based on the decision tree model, realizing the rapid and accurate evaluation of the degree of heart rate abnormality of users. The decision tree model uses real-time collected data for layer-by-layer recursive decision-making, and combines a dynamically updated historical data set to continuously optimize the prediction performance of the model, significantly improving the sensitivity to heart rate change trends and the accuracy of analysis. Through the continuous quantification of the heart rate abnormality index, the present invention can comprehensively and dynamically evaluate the health status of the heart, providing users with refined health monitoring and personalized warning services. In the case of heart rate abnormality being identified, the system can immediately trigger a dynamic warning mechanism and output abnormal information in a timely manner to help users detect potential risks early. The present invention not only greatly improves the intelligence and efficiency of heart rate monitoring, but also provides strong technical support and decision-making basis for the prevention and intervention of heart-related diseases, showing broad application prospects in the field of personalized medicine. Description of the drawings

[0084] Figure 1 It is the specific step flow block diagram of the intelligent heart rate monitoring method based on heart sound and electrocardiogram signals of the present invention;

[0085] Figure 2It is the flowchart of the intelligent heart rate monitoring system based on heart sound and electrocardiogram signals in the present invention. Detailed implementation manners

[0086] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments in the following description are only examples, and those skilled in the art can think of other obvious variations.

[0087] Please refer to Figure 1 As shown, the present invention is an intelligent heart rate monitoring method based on heart sound and electrocardiogram signals, including the following steps:

[0088] Collect the heart sound signal and electrocardiogram signal of the user in real time, evaluate the quality of the collected heart sound signal and electrocardiogram signal, determine the accuracy of the data through the signal quality index, eliminate noise interference and abnormal signals, and extract accurate heart sound signals and electrocardiogram signals;

[0089] Extract the time-domain features and frequency-domain features of the first heart sound and the second heart sound in the heart sound signal, perform fusion processing on the time-domain features and frequency-domain features, and calculate the heart sound abnormality coefficient to evaluate the abnormality degree of the heart sound signal;

[0090] Extract the vibration features of the P wave in the electrocardiogram signal, including: vibration amplitude feature and vibration frequency feature, process the vibration features, and calculate the electrocardiogram abnormality coefficient through comprehensive processing to evaluate the stability of the electrocardiogram signal;

[0091] Based on the heart sound abnormality coefficient and the electrocardiogram abnormality coefficient, dynamically calculate the heart rate abnormality index, divide the user's heart rate into abnormal heart rate and normal heart rate according to the heart rate abnormality index, and trigger a dynamic warning mechanism based on the abnormal heart rate to output heart rate abnormality information.

[0092] Collect the heart sound signal and electrocardiogram signal of the user in real time, evaluate the quality of the collected heart sound signal and electrocardiogram signal, determine the accuracy of the data through the signal quality index, eliminate noise interference and abnormal signals, and extract accurate heart sound signals and electrocardiogram signals, specifically including:

[0093] Obtain signal data through a heart sound sensor and an electrocardiogram electrode, including: heart sound signal and electrocardiogram signal;

[0094] The process of obtaining the signal quality index is: preprocess the collected signal data;

[0095] Obtain the preprocessed signal data;

[0096] Perform further filtering optimization on the heart sound signal and the electrocardiogram signal, dynamically adjust the filter coefficients, obtain the optimized signal data, and calculate the error signal of the signal;

[0097] Calculate the signal-to-noise ratio of the signal data, and the calculation expression is:

[0098]

[0099] In the formula, x(t) represents the preprocessed signal data, and n(t) represents the noise signal. represents the expected value, and t represents time;

[0100] Perform a fast Fourier transform on the optimized signal and calculate the spectral energy. The calculation expression is:

[0101]

[0102] In the formula, P(f) represents the spectral energy, f represents the frequency, i represents the i-th frequency, N represents the maximum number of frequencies, and X(f i ) represents the Fourier transform result at the i-th frequency;

[0103] Calculate the ratio of the signal-to-noise ratio of the signal data to the spectral energy to obtain the signal quality index. The calculation expression is:

[0104]

[0105] In the formula, Q Z represents the signal quality index of the signal data;

[0106] Judge whether the signal quality index is greater than or equal to the preset threshold. If so, record it as accurate heart sound signal and electrocardiogram signal. If not, record it as inaccurate heart sound signal and electrocardiogram signal.

[0107] It should be noted that: The signal quality coefficient is an index used to measure the data quality of heart sound signals and electrocardiogram signals, reflecting the influence degree of noise and interference factors in the signals; it is a comprehensive measurement standard used to judge whether the signal is suitable for further processing and analysis.

[0108] Extract the time-domain features and frequency-domain features of the first heart sound and the second heart sound in the heart sound signal, perform fusion processing on the time-domain features and frequency-domain features, and calculate the heart sound abnormality coefficient to evaluate the abnormality degree of the heart sound signal. Specifically, it includes:

[0109] Perform denoising processing on the collected heart sound signal. Adopt the wavelet transform denoising algorithm, use the threshold denoising method to remove high-frequency noise and retain the effective heart sound components to obtain the denoised heart sound signal;

[0110] Perform wavelet transform on the denoised heart sound signal to obtain the decomposed wavelet coefficients, and extract the time-domain features of the first heart sound and the second heart sound from the decomposed wavelet coefficients, including the duration, which is calculated through the local extreme values of the wavelet coefficients;

[0111] Among them, the calculation expression of the wavelet transform is as follows:

[0112]

[0113] In the formula, ψ j (t) represents the wavelet basis of the j-th layer, x1(t) represents the heart sound signal, j represents the number of decomposition layers, J represents the total number of decomposition layers, c j represents the wavelet coefficient of the j-th layer, and t represents time;

[0114] In the frequency-domain analysis of the wavelet coefficients, the frequency-domain characteristics of the heart sound signal are extracted, including: the main frequency and the bandwidth; among them, through the spectrum analysis of the wavelet coefficients, the main frequencies of the first heart sound and the second heart sound, and the corresponding bandwidths are calculated;

[0115] Among them, the calculation expression of the main frequency is as follows:

[0116] f peak = argmax f |X(f)|;

[0117] In the formula, f peak represents the main frequency of the heart sound signal, and X(f) is the spectrum of the heart sound signal;

[0118] The bandwidth is obtained by calculating the energy distribution in different frequency bands;

[0119] The extracted time-domain characteristics and frequency-domain characteristics are fused, and the comprehensive feature vector is calculated. The calculation expression is as follows:

[0120] F fusion = α·F time + β·F freq ;

[0121] In the formula, F fusion represents the comprehensive feature vector, F time represents the time-domain feature vector, F freq represents the frequency-domain feature vector, and α and β are preset proportional coefficients;

[0122] According to the fused features, the heart sound abnormality coefficient is calculated. The calculation expression is as follows:

[0123]

[0124] In the formula, EC represents the heart sound abnormality coefficient, F normal represents the feature vector of the normal heart sound, and σ normal represents the standard deviation of the normal heart sound feature;

[0125] Determine whether the abnormal coefficient of heart sound is greater than or equal to a preset threshold. If so, the abnormal degree of the heart sound signal is high; if not, the abnormal degree of the heart sound signal is low.

[0126] It should be noted that: the abnormal coefficient of heart sound is an index used to evaluate the abnormal degree of the heart sound signal; the heart sound signal contains two main heart sounds: the first heart sound and the second heart sound, which are generated during the contraction and relaxation of the heart; if abnormal heart sounds (such as murmurs or irregularities) appear in the signal, the abnormal coefficient of heart sound can be used to quantify these abnormal characteristics.

[0127] Extract the vibration characteristics of the P wave in the electrocardiogram signal, including: vibration amplitude characteristics and vibration frequency characteristics, process the vibration characteristics, and calculate the electrocardiogram abnormal coefficient through comprehensive processing for evaluating the stability of the electrocardiogram signal, specifically including:

[0128] The processing of the vibration characteristics specifically includes:

[0129] Extract the vibration amplitude characteristics and vibration frequency characteristics of the P wave in the electrocardiogram signal, calculate the vibration amplitude abnormal coefficient according to the vibration amplitude characteristics, and calculate the vibration frequency fluctuation coefficient according to the fluctuation degree of the vibration frequency characteristics;

[0130] Among them, the acquisition process of the vibration amplitude abnormal coefficient is:

[0131] Extract the P wave signal from the electrocardiogram signal;

[0132] Perform vibration amplitude calculation on the extracted P wave signal to obtain the vibration amplitude, and the calculation expression is:

[0133] A P =max(x Pwave (t))-min(x Pwave (t));

[0134] In the formula, A P represents the vibration amplitude, x Pwave (t) represents the P wave signal, and t represents time;

[0135] Based on a group of normal P wave signals, calculate the mean and standard deviation of the normal vibration amplitude, and then calculate the vibration amplitude abnormal coefficient of the current P wave, and the calculation expression is:

[0136]

[0137] In the formula, EC A represents the vibration amplitude abnormal coefficient, μ normal represents the mean of the vibration amplitude, and σ normal represents the standard deviation of the vibration amplitude;

[0138] Among them, the process of obtaining the vibration frequency fluctuation coefficient is as follows:

[0139] Extract the P-wave signal from the electrocardiogram (ECG) signal;

[0140] Use the fast Fourier transform to perform frequency-domain analysis on the P-wave signal and extract the main frequency of the P-wave;

[0141] Calculate the fluctuation amplitude of the P-wave frequency. The calculation expression is:

[0142] Δf Pwave =max(f Pwave (t)) - min(f Pwave (t));

[0143] In the formula, Δf Pwave represents the fluctuation amplitude of the P-wave frequency, and f Pwave (t) represents the P-wave frequency extracted at different time points;

[0144] Based on the standard deviation of the P-wave frequency, calculate the vibration frequency fluctuation coefficient. The calculation expression is:

[0145]

[0146] In the formula, EC f represents the vibration frequency fluctuation coefficient, and σ f represents the standard deviation of the P-wave frequency;

[0147] Obtain the vibration amplitude abnormality coefficient and the vibration frequency fluctuation coefficient of the P-wave in the ECG signal, and perform normalization processing on the vibration amplitude abnormality coefficient and the vibration frequency fluctuation coefficient to calculate the ECG abnormality coefficient;

[0148] The calculation expression of the ECG abnormality coefficient is:

[0149]

[0150] In the formula, EC d represents the ECG abnormality coefficient, a1 and a2 are preset proportionality coefficients, and both a1 and a2 are greater than 0, and e represents the natural logarithm coefficient;

[0151] Judge whether the ECG abnormality coefficient is greater than or equal to the preset threshold. If so, the degree of abnormality in the ECG signal is high. If not, the degree of abnormality in the ECG signal is low.

[0152] It should be noted that: The ECG abnormality coefficient is used to evaluate the degree of abnormality in the ECG signal, and when the value of the ECG abnormality coefficient is larger, the degree of abnormality in the corresponding user's ECG signal is higher.

[0153] Based on the heart sound anomaly coefficient and the electrocardiogram anomaly coefficient, dynamically calculate the heart rate anomaly index. According to the heart rate anomaly index, divide the user's heart rate into abnormal heart rate and normal heart rate, and based on the abnormal heart rate, trigger a dynamic warning mechanism to output heart rate anomaly information, specifically including:

[0154] Collect the user's heart sound signal and electrocardiogram signal in real time, and calculate and obtain the heart sound anomaly coefficient and the electrocardiogram anomaly coefficient respectively;

[0155] Construct a decision tree model based on the historical data set, and the historical data set includes the heart sound anomaly coefficient, the electrocardiogram anomaly coefficient, and the corresponding heart rate anomaly index label;

[0156] The decision tree selects the heart sound anomaly coefficient and the electrocardiogram anomaly coefficient as features, selects the splitting point based on the splitting criterion, and recursively constructs the decision tree model;

[0157] For the newly collected heart sound anomaly coefficient and electrocardiogram anomaly coefficient in real time, perform recursive decision-making according to the decision tree model through the layer-by-layer judgment rule;

[0158] Output the heart rate anomaly index according to the recursive decision, and the heart rate anomaly index is a continuous value;

[0159] By dynamically inputting new heart sound and electrocardiogram signal data, update the decision tree model in real time to further improve the accuracy and real-time performance of the calculation results;

[0160] Calculate the mean value of the heart rate anomaly index within a period of time monitored by the user, and judge whether the heart rate anomaly index within the current time period is greater than or equal to the preset threshold. If so, the user's heart rate within the corresponding time period is abnormal and is recorded as an abnormal heart rate. If not, the user's heart rate within the corresponding time period is normal and is recorded as a normal heart rate;

[0161] It should be noted that: The heart rate anomaly index reflects the overall health status of the heart, evaluates whether the heart rate is normal, and the greater the value of the heart rate anomaly index, the higher the degree of abnormality of the overall health status of the corresponding heart.

[0162] Please refer to Figure 2 As shown, the intelligent heart rate monitoring system based on heart sound and electrocardiogram signals includes:

[0163] A data acquisition module that collects the user's heart sound signal and electrocardiogram signal in real time;

[0164] A signal quality evaluation module that evaluates the quality of the collected heart sound signal and electrocardiogram signal, determines the accuracy of the data through the signal quality index, eliminates noise interference and abnormal signals, and extracts accurate heart sound signals and electrocardiogram signals;

[0165] A heart sound signal evaluation module, which extracts the time-domain features and frequency-domain features of the first heart sound and the second heart sound in the heart sound signal, performs fusion processing on the time-domain features and frequency-domain features, calculates the heart sound abnormality coefficient, and is used to evaluate the abnormality degree of the heart sound signal;

[0166] An electrocardiogram signal evaluation module, which extracts the vibration features of the P wave in the electrocardiogram signal, including: vibration amplitude feature and vibration frequency feature, processes the vibration features, and calculates the electrocardiogram abnormality coefficient through comprehensive processing, and is used to evaluate the stability of the electrocardiogram signal;

[0167] A heart rate abnormality dynamic judgment module, which dynamically calculates the heart rate abnormality index based on the heart sound abnormality coefficient and the electrocardiogram abnormality coefficient, divides the user's heart rate into abnormal heart rate and normal heart rate according to the heart rate abnormality index, and triggers a dynamic warning mechanism based on the abnormal heart rate to output heart rate abnormality information.

[0168] The working principle of the present invention: The heart sound signal and the electrocardiogram signal of the user are collected in real time through a heart sound sensor and an electrocardiogram electrode, the signal quality index is used to evaluate the quality of the collected signal data, noise interference and abnormal signals are removed, and accurate heart sound signals and electrocardiogram signals are extracted; the time-domain features and frequency-domain features of the first heart sound and the second heart sound in the heart sound signal are extracted, and the heart sound abnormality coefficient is calculated after fusion processing to evaluate the abnormality degree of the heart sound signal; the P wave vibration features in the electrocardiogram signal are extracted, including the vibration amplitude feature and the vibration frequency feature, and the electrocardiogram abnormality coefficient is calculated based on the vibration features to evaluate the stability of the electrocardiogram signal; based on the heart sound abnormality coefficient and the electrocardiogram abnormality coefficient, a decision tree model is constructed, and the heart rate abnormality index is dynamically calculated through the model to evaluate the abnormality degree of the heart rate, the user's heart rate is divided into abnormal heart rate and normal heart rate, and a dynamic warning mechanism is triggered to output heart rate abnormality information. The present invention realizes the accurate monitoring and real-time warning of heart rate abnormality through the combination of multi-parameter fusion analysis and intelligent algorithm, and provides an efficient and reliable solution for the continuous monitoring of heart health.

[0169] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by software simulation of a large amount of collected data to get a formula closest to the actual situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0170] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wire or wirelessly (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that contains one or more collections of available media. The available media can be magnetic media (such as floppy disks, hard disks, magnetic tapes), optical media (such as DVDs), or semiconductor media. The semiconductor media can be a solid-state drive.

[0171] It should be understood that the term "and / or" in this document is merely a description of the association relationship between associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. Here, A and B can be singular or plural. In addition, the character " / " in this document generally represents an "or" relationship between the associated objects before and after, but it may also represent an "and / or" relationship, which can be specifically understood by referring to the context before and after.

[0172] It should be understood that in various embodiments of the present application, the magnitudes of the serial numbers of the above processes do not mean the order of execution. The order of execution of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.

[0173] The above has described in detail one embodiment of the present invention, but the content described is only a preferred embodiment of the present invention and cannot be considered as limiting the scope of implementation of the present invention. Any equivalent changes and improvements made within the scope of the application of the present invention should still fall within the scope covered by the patent of the present invention.

Claims

1. An intelligent heart rate monitoring method based on heart sound and electrocardiogram signals, characterized in that: The following steps are involved: Collect the user's heart sound signals and ECG signals in real time, evaluate the quality of the collected heart sound signals and ECG signals, determine the accuracy of the data through the signal quality index, eliminate noise interference and abnormal signals, and extract accurate heart sound signals and ECG signals; Extract the time domain features and frequency domain features of the first heart sound and the second heart sound in the heart sound signal, fuse the time domain features and the frequency domain features, calculate the heart sound abnormality coefficient, and use it to evaluate the abnormality of the heart sound signal; Extract the vibration characteristics of the P wave in the ECG signal, including: vibration amplitude characteristics and vibration frequency characteristics, process the vibration characteristics, and calculate the ECG abnormality coefficient through comprehensive processing to evaluate the stability of the ECG signal; Based on the abnormal heart sound coefficient and the abnormal electrocardiogram coefficient, the heart rate abnormality index is dynamically calculated. According to the abnormal heart rate index, the user's heart rate is divided into abnormal heart rate and normal heart rate. Based on the abnormal heart rate, the dynamic early warning mechanism is triggered to output abnormal heart rate information.

2. The intelligent heart rate monitoring method based on heart sound and electrocardiogram signals according to claim 1, characterized in that: The process of obtaining the signal quality index is as follows: Acquire signal data through the heart sound sensor and the electrocardiogram electrode, including: heart sound signal and electrocardiogram signal; Preprocessing the collected signal data; Obtain preprocessed signal data; Further filter and optimize the heart sound signal and the ECG signal, dynamically adjust the filter coefficient, obtain the optimized signal data, and calculate the error signal of the signal; Calculate the signal-to-noise ratio of the signal data. The calculation expression is: In the formula, x(t) represents the preprocessed signal data, n(t) represents the noise signal, represents the expected value, t represents the time; Perform fast Fourier transform on the optimized signal and calculate the spectrum energy. The calculation expression is: Where P(f) represents the spectrum energy, f represents the frequency, i represents the i-th frequency, N represents the maximum number of frequencies, and X(f i ) represents the Fourier transform result at the i-th frequency; The signal quality index is obtained by calculating the ratio of the signal-to-noise ratio of the signal data to the spectrum energy. The calculation expression is: In the formula, Q Z A signal quality index representing the signal data.

3. The intelligent heart rate monitoring method based on heart sound and electrocardiogram signals according to claim 1, characterized in that: The extracting of accurate heart sound signals and electrocardiogram signals specifically includes: It is determined whether the signal quality index is greater than or equal to a preset threshold. If so, it is recorded as an accurate heart sound signal and electrocardiogram signal. If not, it is recorded as an inaccurate heart sound signal and electrocardiogram signal.

4. The intelligent heart rate monitoring method based on heart sound and electrocardiogram signals according to claim 1, characterized in that: The process of obtaining the abnormal heart sound coefficient is as follows: The collected heart sound signals are denoised by using a wavelet transform denoising algorithm and a threshold denoising method to remove high-frequency noise and retain effective heart sound components to obtain denoised heart sound signals; Performing wavelet transform on the denoised heart sound signal to obtain decomposed wavelet coefficients, extracting the time domain features of the first heart sound and the second heart sound from the decomposed wavelet coefficients, including the duration, by calculating the local extreme values ​​of the wavelet coefficients; The calculation expression of the wavelet transform is: In the formula, ψ j (t) represents the j-th wavelet basis, x1(t) represents the heart sound signal, j represents the number of decomposition layers, J represents the total number of decomposition layers, c j represents the j-th layer wavelet coefficient, and t represents time; In the frequency domain analysis of the wavelet coefficients, the frequency domain features of the heart sound signal are extracted, including: the main frequency and the bandwidth; wherein, the main frequencies of the first heart sound and the second heart sound, and the corresponding bandwidths are calculated through the spectrum analysis of the wavelet coefficients; Wherein, the calculation expression of the main frequency is: f peak =argmax f |X(f)|; In the formula, f peak represents the main frequency of the heart sound signal, X(f) is the spectrum of the heart sound signal; The bandwidth is obtained by calculating the energy distribution in different frequency bands; The extracted time domain features and frequency domain features are fused to calculate the comprehensive feature vector, and the calculation expression is: F fusion =α·F time +β·F freq ; In the formula, F fusion represents the comprehensive feature vector, F time represents the time domain feature vector, F freq represents the frequency domain feature vector, α and β are preset proportional coefficients; The heart sound abnormality coefficient is calculated based on the fused features. The calculation expression is: In the formula, EC represents the coefficient of abnormal heart sound, F normal The eigenvector representing normal heart sounds, σ normal Represents the standard deviation of normal heart sound characteristics.

5. The intelligent heart rate monitoring method based on heart sound and electrocardiogram signals according to claim 1, characterized in that: The processing of the vibration characteristics specifically includes: Extract the vibration amplitude characteristics and vibration frequency characteristics of the P wave in the electrocardiogram signal, calculate the vibration amplitude abnormality coefficient according to the vibration amplitude characteristics, and calculate the vibration frequency fluctuation coefficient according to the fluctuation degree of the vibration frequency characteristics; The process of obtaining the vibration amplitude abnormality coefficient is as follows: Extract P wave signal from ECG signal; The vibration amplitude of the extracted P wave signal is calculated to obtain the vibration amplitude. The calculation expression is: A P =max(x Pwave (t))-min(x Pwave (t)); In the formula, A P represents the vibration amplitude, x Pwave (t) represents the P wave signal, and t represents time; Based on a set of normal P wave signals, the mean and standard deviation of the normal vibration amplitude are calculated, and then the abnormal coefficient of the vibration amplitude of the current P wave is calculated. The calculation expression is: In the formula, EC A Indicates the vibration amplitude abnormal coefficient, μ normal represents the mean value of the vibration amplitude, σ normal represents the standard deviation of the vibration amplitude; The process of obtaining the vibration frequency fluctuation coefficient is as follows: Extract P wave signal from ECG signal; Use fast Fourier transform to perform frequency domain analysis on the P wave signal and extract the main frequency of the P wave; Calculate the fluctuation amplitude of the P wave frequency, the calculation expression is: Δf Pwave =max(f Pwave (t))-min(f Pwave (t)); Where Δf Pwave represents the fluctuation amplitude of the P wave frequency, f Pwave (t) represents the P wave frequency extracted at different time points; Based on the standard deviation of the P-wave frequency, the vibration frequency fluctuation coefficient is calculated, and the calculation expression is: In the formula, EC f represents the vibration frequency fluctuation coefficient, σ f represents the standard deviation of the P wave frequency.

6. The intelligent heart rate monitoring method based on heart sound and electrocardiogram signals according to claim 1, characterized in that: The acquisition process of the ECG abnormality coefficient is as follows: The vibration amplitude abnormality coefficient and the vibration frequency fluctuation coefficient of the P wave in the electrocardiogram signal are obtained, and the vibration amplitude abnormality coefficient and the vibration frequency fluctuation coefficient are normalized to calculate the electrocardiogram abnormality coefficient.

7. The intelligent heart rate monitoring method based on heart sound and electrocardiogram signals according to claim 1, characterized in that: The method of dynamically calculating the abnormal heart rate index based on the abnormal heart sound coefficient and the abnormal electrocardiogram coefficient specifically includes: Collect the user's heart sound signal and electrocardiogram signal in real time, and calculate and obtain the heart sound abnormality coefficient and electrocardiogram abnormality coefficient respectively; Building a decision tree model based on a historical data set, wherein the historical data set includes an abnormal heart sound coefficient, an abnormal electrocardiogram coefficient, and a corresponding abnormal heart rate index label; The decision tree selects the heart sound abnormality coefficient and the electrocardiogram abnormality coefficient as features, selects the segmentation point based on the splitting criterion, and recursively constructs a decision tree model; For the newly acquired abnormal heart sound coefficients and abnormal electrocardiogram coefficients in real time, recursive decisions are made through layer-by-layer judgment rules according to the decision tree model; Outputting the abnormal heart rate index according to recursive decision making, wherein the abnormal heart rate index is a continuous value; By dynamically inputting new heart sound and ECG signal data, the decision tree model is updated in real time to further improve the accuracy and real-time performance of the calculation results.

8. The intelligent heart rate monitoring method based on heart sound and electrocardiogram signals according to claim 1, characterized in that: The method of classifying the user's heart rate into abnormal heart rate and normal heart rate according to the abnormal heart rate index specifically includes: Calculate the average of the abnormal heart rate index during a period of user monitoring, and determine whether the abnormal heart rate index in the current time period is greater than or equal to the preset threshold. If so, the user's heart rate in the corresponding time period is abnormal and recorded as an abnormal heart rate. If not, the user's heart rate in the corresponding time period is normal and recorded as a normal heart rate.

9. An intelligent heart rate monitoring system based on heart sound and electrocardiogram signals, characterized in that: The intelligent heart rate monitoring method based on heart sound and electrocardiogram signals as claimed in any one of claims 1 to 8 comprises: A data acquisition module, which collects the user's heart sound signals and electrocardiogram signals in real time; A signal quality evaluation module, which evaluates the quality of the collected heart sound signals and electrocardiogram signals, determines the accuracy of the data through the signal quality index, removes noise interference and abnormal signals, and extracts accurate heart sound signals and electrocardiogram signals; A heart sound signal evaluation module, wherein the heart sound signal evaluation module extracts time domain features and frequency domain features of the first heart sound and the second heart sound in the heart sound signal, fuses the time domain features and the frequency domain features, and calculates a heart sound abnormality coefficient for evaluating the abnormality of the heart sound signal; An ECG signal evaluation module, wherein the ECG signal evaluation module extracts vibration characteristics of the P wave in the ECG signal, including vibration amplitude characteristics and vibration frequency characteristics, processes the vibration characteristics, and calculates an ECG abnormality coefficient through comprehensive processing to evaluate the stability of the ECG signal; The abnormal heart rate dynamic judgment module dynamically calculates the abnormal heart rate index based on the abnormal heart sound coefficient and the abnormal electrocardiogram coefficient, divides the user's heart rate into abnormal heart rate and normal heart rate according to the abnormal heart rate index, and triggers a dynamic early warning mechanism based on the abnormal heart rate to output abnormal heart rate information.

Images