A risk prediction method, device and equipment based on heart sound electrocardio and medium

Abstract

The present application relates to the medical health technical field, especially a risk prediction method, device, equipment and medium based on heart sound electrocardio, synchronous acquisition aortic valve, mitral valve and tricuspid valve three key auscultation area heart sound signal and corresponding electrocardio signal, form multidimensional physiological information source, adopt preset time window to four way signal synchronous segmentation, ensure each signal segment is strictly aligned in time, lay a foundation for subsequent fusion analysis. In the signal fusion stage, three sub heart sound signals are weighted and fused based on electrocardio gating and cross valve area coherence weight, effectively suppress the noise interference and pathological information missing problem that may exist in single auscultation area signal, based on the fused heart sound signal and sub electrocardio signal, the evidence abnormality measure of heart sound signal is calculated, the risk prediction result is obtained according to the evidence abnormality measure of the measured person, without additional dependence on complex priori hypothesis or artificial intervention, ensure the timeliness of risk early warning.

Description

Technical Field

[0001] This invention relates to the field of medical and health technology, and in particular to a risk prediction method, device, equipment and medium based on heart sounds and electrocardiograms. Background Technology

[0002] With the continued aging of the global population and changes in lifestyle, cardiovascular diseases such as heart failure have become prevalent chronic end-stage diseases among the elderly, with their incidence and disability rates rising year by year, seriously threatening public health and safety. Currently, clinical prediction of heart failure mainly relies on electrocardiography, echocardiography, and biomarkers. While these methods have high accuracy, they are highly dependent on medical institutions and professionals, making it difficult to provide timely warnings for heart failure patients. Therefore, how to conduct timely predictions for cardiovascular diseases such as heart failure has become an urgent problem to be solved. Summary of the Invention

[0003] In view of this, embodiments of this application provide a risk prediction method, device, equipment, and medium based on heart sounds and electrocardiograms to solve the problem of untimely prediction in the prediction of cardiovascular diseases such as heart failure.

[0004] In a first aspect, embodiments of this application provide a risk prediction method based on heart sounds and electrocardiograms, the risk prediction method comprising: Acquire the heart sound signals and electrocardiogram signals of the subject to be tested. The heart sound signals include the first heart sound signal in the aortic valve auscultation area, the second heart sound signal in the mitral valve auscultation area, and the third heart sound signal in the tricuspid valve auscultation area. According to a preset time window, the first channel heart sound signal, the second channel heart sound signal, the third channel heart sound signal and the electrocardiogram signal are segmented to obtain multiple segments of first sub-heart sound signals of the first channel heart sound signal, multiple segments of second sub-heart sound signals of the second channel heart sound signal, multiple segments of third sub-heart sound signals of the third channel heart sound signal and multiple segments of electrocardiogram signals of the electrocardiogram signal; For the first sub-heart sound signal, the second sub-heart sound signal, the third sub-heart sound signal, and the sub-electrocardiogram signal of the corresponding segment within any time window, the first sub-heart sound signal, the second sub-heart sound signal, and the third sub-heart sound signal are fused according to the sub-electrocardiogram signal to obtain the fused heart sound signal; The evidence of abnormality in the heart sound signal is calculated by fusing the heart sound signal with the sub-electrocardiogram signal within the corresponding segment. Risk prediction is performed on the individuals to be tested based on the evidence anomaly measure, and the risk prediction results are obtained.

[0005] Secondly, this application provides a risk prediction device based on heart sound and electrocardiogram. The risk prediction device includes a main control module, a three-lead electrode sheet and multiple electret microphones. The three-lead electrode sheet is connected to the main control module through a retractable lead wire, and each electret microphone is connected to the main control module through a flexible retractable connection structure. The three-lead electrode pads are used to collect the electrocardiogram (ECG) signals of the person being tested and transmit the ECG signals to the main control module; Multiple electret microphones are used to collect the first heart sound signal from the aortic valve auscultation area, the second heart sound signal from the mitral valve auscultation area, and the third heart sound signal from the tricuspid valve auscultation area, and transmit the first, second, and third heart sound signals to the main control module; The main control module is used to execute any of the above-mentioned risk prediction methods based on heart sounds and electrocardiograms.

[0006] Thirdly, embodiments of this application provide a computer device, which includes a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the risk prediction method described above.

[0007] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the risk prediction method described above.

[0008] The advantages of this application compared to the prior art are: This application simultaneously acquires heart sound signals and corresponding electrocardiogram (ECG) signals from three key auscultation areas: the aortic, mitral, and tricuspid valves, forming a multidimensional physiological information source. A preset time window is used to synchronously segment the four signals, ensuring strict temporal alignment and laying the foundation for subsequent fusion analysis. During the signal fusion stage, the three sub-heart sound signals are weighted and fused based on ECG gating and cross-valve region coherence weights. This effectively suppresses noise interference and missing pathological information that may exist in signals from a single auscultation area. Based on the fused heart sound signals and sub-ECG signals, an evidence anomalous measure is calculated. Risk prediction is then performed on the individuals being tested based on this evidence anomalous measure, yielding risk prediction results without relying on complex prior assumptions or manual intervention, ensuring timely risk warning. Attached Figure Description

[0009] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 This is a flowchart illustrating a risk prediction method based on heart sounds and electrocardiograms provided in an embodiment of this application; Figure 2 This is a schematic diagram of the structure of a risk prediction device based on heart sound and electrocardiogram according to an embodiment of this application; Figure 3 This is another structural schematic diagram of a risk prediction device based on heart sound and electrocardiogram provided in one embodiment of this application; Figure 4 This is a schematic diagram of the structure of a risk prediction device based on heart sound and electrocardiogram provided in an embodiment of this application from different perspectives; Figure 5 This is a schematic diagram of wearing a risk prediction device based on heart sound and electrocardiogram according to an embodiment of this application; Figure 6 This is a schematic diagram of the structure of a computer device provided in one embodiment of this application; The following are the labeling elements in the figure: Main control module 400, three-lead electrode pads 100, electret microphone 300, retractable lead wire 200, flexible telescopic connection structure 310, ECG lead wire telescopic module 500, telescopic handle 510, and reel 520. Detailed Implementation

[0011] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0012] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0013] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0014] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0015] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0016] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0017] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0018] It should be understood that the sequence number of each step in the following embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0019] To illustrate the technical solution of this application, specific embodiments are described below.

[0020] like Figure 1 As shown, Figure 1 This is a flowchart illustrating a risk prediction method based on heart sound and electrocardiogram according to an embodiment of this application, as shown below. Figure 1 As shown, the risk prediction method based on heart sounds and electrocardiograms may include the following steps.

[0021] S101: Acquire the heart sound and electrocardiogram signals of the subject. The heart sound signals include the first heart sound signal from the aortic valve auscultation area, the second heart sound signal from the mitral valve auscultation area, and the third heart sound signal from the tricuspid valve auscultation area.

[0022] In step S101, heart sound signals are acquired using electret microphones placed in the aortic, mitral, and tricuspid auscultation areas of the subject's chest. Electrocardiogram (ECG) signals are acquired using three-lead electrode pads placed at corresponding lead positions on the subject's body surface.

[0023] In this embodiment, the heart sound acquisition module collects the heart sound signals of the subject. This module includes three electret microphones, placed in the aortic valve auscultation area, mitral valve auscultation area, and tricuspid valve auscultation area, respectively. The heart sound signals include a first signal from the aortic valve auscultation area, a second signal from the mitral valve auscultation area, and a third signal from the tricuspid valve auscultation area. This multi-point signal acquisition effectively improves the quality and accuracy of the heart sound signal acquisition.

[0024] The ECG acquisition module collects the ECG signals of the subject. This module includes a three-lead electrode pad, which is a common circular electrode pad found in the market. The base of the three-lead electrode pad is made of non-woven fabric with ventilation holes, providing a larger contact area while ensuring comfort and breathability during prolonged wear. The outer part of the three-lead electrode pad is coated with a bio-gel to ensure stable contact with the skin surface, reducing contact impedance fluctuations and the risk of electrode detachment caused by skin laxity, sensitivity, or sweat, and minimizing artifacts that interfere with subsequent abnormal signal identification and risk warning judgment.

[0025] In this embodiment, by simultaneously acquiring heart sound signals from multiple regions, the sound characteristics generated by the activity of different heart valves can be comprehensively captured, avoiding signal omissions or distortions that may occur with a single auscultation area. Simultaneously, combined with electrocardiogram (ECG) signals, this provides a data foundation for the accurate identification of subsequent abnormal signals and reliable risk prediction.

[0026] S102: The first heart sound signal, the second heart sound signal, the third heart sound signal and the electrocardiogram signal are segmented according to the preset time window to obtain multiple segments of the first sub-heart sound signal of the first heart sound signal, multiple segments of the second sub-heart sound signal of the second heart sound signal, multiple segments of the third sub-heart sound signal of the third heart sound signal and multiple segments of the electrocardiogram signal.

[0027] In step S102, the preset time window is a fixed duration that is set in advance, that is, the time length of the window.

[0028] In this embodiment, the first, second, and third heart sound signals, along with the electrocardiogram (ECG) signal, are segmented according to a preset time window to obtain multiple segments of first sub-heart sound signals from the first heart sound signal, multiple segments of second sub-heart sound signals from the second heart sound signal, multiple segments of third sub-heart sound signals from the third heart sound signal, and multiple segments of ECG signals from the ECG signal. For example, the duration is set to 5 seconds, and the step size is set to 2 seconds, meaning that a 5-second signal segment is extracted every 2 seconds starting from the signal start time. During the segmentation process, edge segments with a duration of less than 5 seconds at the beginning and end of the signal are padded with zeros to ensure the consistency of the length of each sub-signal segment.

[0029] Optionally, the first, second, and third heart sound signals and the electrocardiogram (ECG) signal are segmented according to a preset time window to obtain multiple segments of first sub-heart sound signals from the first heart sound signal, multiple segments of second sub-heart sound signals from the second heart sound signal, multiple segments of third sub-heart sound signals from the third heart sound signal, and multiple segments of sub-ECG signals from the ECG signal, including: The first, second, and third heart sound signals, along with the electrocardiogram (ECG) signal, are converted into signal sequences on a unified sampling grid, resulting in the first resampled heart sound signal sequence of the first heart sound signal, the second resampled heart sound signal sequence of the second heart sound signal, the third resampled heart sound signal sequence of the third heart sound signal, and the resampled ECG signal sequence of the ECG signal. According to the preset time window, the first resampled heart sound signal sequence, the second resampled heart sound signal sequence, the third resampled heart sound signal sequence and the resampled electrocardiogram signal sequence are segmented to obtain multiple segments of the first sub-heart sound signal of the first heart sound signal, multiple segments of the second sub-heart sound signal of the second heart sound signal, multiple segments of the third sub-heart sound signal of the third heart sound signal and multiple segments of the electrocardiogram signal.

[0030] In this embodiment, when converting the first, second, and third heart sound signals and the electrocardiogram (ECG) signal into a signal sequence on a unified sampling grid, the absolute timestamps of different sampling points of the first, second, and third heart sound signals and the ECG signal on the same time axis are calculated. The first clock count value of the first, second, and third heart sound signals and the second clock count value of the ECG signal are obtained. The first clock count value is the clock count value of the corresponding sampling point for each heart sound signal after time delay compensation. Both the first and second clock count values ​​are clock count values ​​under the MCU system clock.

[0031] The initial clock count and clock frequency are obtained. The initial clock count is the clock count value corresponding to the initial moment when the MCU system clock starts counting during the acquisition of heart sound and ECG signals. It serves as the reference for the clock count values ​​of all subsequent sampling points. The clock frequency refers to the operating frequency of the MCU system clock, usually measured in Hertz (Hz). Based on the first clock count, the initial clock count, and the clock frequency, the first absolute timestamp of the heart sound signal sampling point is calculated. Based on the second clock count, the initial clock count, and the clock frequency, the second absolute timestamp of the ECG signal sampling point is calculated. The calculation formula is as follows: in, For the heart sound signal The first absolute timestamp of the sampling point, i.e., the first, second, and third heart sound signals. The first absolute timestamp of each sampling point For ECG signal number The second absolute timestamp of each sampling point For the first The count value of each heart sound signal sampling point under the MCU system clock. This is the initial clock count value, which is the trigger moment when the heart sound signal and electrocardiogram signal are acquired synchronously. For clock frequency, For the first The count values ​​of each ECG signal sampling point under the MCU system clock. Among them, the first absolute timestamp and the second absolute timestamp are timestamps on the same time axis.

[0032] Based on the first and second absolute timestamps, the corresponding heart sound signals and electrocardiogram (ECG) signals are cross-modal time-aligned to obtain sequences on a unified sampling grid. Specifically, the heart sound signals and ECG signals on the same time axis are obtained based on the first and second absolute timestamps. A target sampling rate is set for any one of the first, second, and third heart sound signals and the ECG signal. Reconstructing the nucleus using multiphase reconstruction For the original sequence Make the insertion-to-pulling ratio as The uniform reconstruction yields a signal sequence on a uniform sampling grid. : in, The output sequence after resampling, i.e., the signal sequence on the uniform sampling grid, can be the first resampled heart sound signal sequence of the first channel heart sound signal, the second resampled heart sound signal sequence of the second channel heart sound signal, the third resampled heart sound signal sequence of the third channel heart sound signal, or the resampled ECG signal sequence of the ECG signal. n is the new sample index, i.e., the sampling point index on the uniform sampling grid. The original signal sequence includes the first heart sound signal, the second heart sound signal, the third heart sound signal, and the electrocardiogram signal. k is the original sample index, i.e., the sampling point index in the original signal sequence. This is a low-pass prototype of a multiphase FIR filter.

[0033] The first, second, and third resampled heart sound signal sequences, along with the resampled electrocardiogram (ECG) signal sequence, are segmented according to a preset time window to obtain multiple segments of first sub-heart sound signals from the first heart sound signal, multiple segments of second sub-heart sound signals from the second heart sound signal, multiple segments of third sub-heart sound signals from the third heart sound signal, and multiple segments of sub-ECG signals from the ECG signal. The frame length and step size are calculated based on the duration of the preset time window and the sampling rate, where the sampling rate is the number of sampling points per second set during the acquisition of the heart sound and ECG signals. If the sampling rates of the heart sound and ECG signals are different, rational number decimation or interpolation can be used to achieve a unified sampling rate. The formulas for calculating the frame length and step size are as follows: in, The frame length is the number of sampling points within a preset time window. The duration of the preset time window, Sampling rate, To round down and ensure the frame length is an integer number of sampling points, Step size, The overlap ratio is approximately 0.5, meaning that adjacent segments overlap by about 50%.

[0034] Based on the corresponding frame length and step size, determine the first frame in the resampled output sequence. Segment starting sampling point According to the Segment starting sampling point Calculate the first The start and end timestamps of the segment. The calculation formula is as follows: in, For the first The start timestamp of the segment. For the first End timestamp of the segment To unify the sampling grid The count value of the segment's starting sampling point under the MCU system clock. This is the initial clock count value. For clock frequency, For frame length, The sampling rate.

[0035] Based on the start and end timestamps of each segment, the first resampled heart sound signal sequence, the second resampled heart sound signal sequence, the third resampled heart sound signal sequence, and the resampled electrocardiogram (ECG) signal sequence are segmented to obtain multiple segments of first sub-heart sound signals from the first heart sound signal, multiple segments of second sub-heart sound signals from the second heart sound signal, multiple segments of third sub-heart sound signals from the third heart sound signal, and multiple segments of sub-ECG signals from the ECG signal. Within any given time window, the first, second, third, and sub-ECG signals of the corresponding segment are signal sequences on a uniform sampling grid, i.e., the resampled output sequence.

[0036] In this embodiment, the segmentation processing method based on absolute timestamps described above can accurately align multiple heart sound signals and electrocardiogram signals from different acquisition channels onto the same time axis, ensuring the synchronicity of each signal in the time dimension. This effectively avoids feature extraction deviations caused by signal asynchrony.

[0037] S103: For the first sub-heart sound signal, the second sub-heart sound signal, the third sub-heart sound signal and the sub-electrocardiogram signal of the corresponding segment within any time window, the first sub-heart sound signal, the second sub-heart sound signal and the third sub-heart sound signal are fused according to the sub-electrocardiogram signal to obtain the fused heart sound signal.

[0038] In step S103, the sub-heart sound signals in each path within the same time window are fused to obtain a fused heart sound signal.

[0039] In this embodiment, for the first, second, and third sub-heart sound signals and the sub-ECG signal within any given time window, the first, second, and third sub-heart sound signals are fused based on the sub-ECG signal to obtain a fused heart sound signal. A time alignment and weighted fusion strategy based on the R-wave position of the ECG signal can be used to achieve effective fusion of multiple heart sound signals. For example, the R-wave peak position is detected from the sub-ECG signal, and the physiological time window for the first heart sound S1 and the second heart sound S2 in the heart sound signal is determined based on the R-wave. Since the ECG signal and the heart sound signal have a clear physiological correspondence, the time interval from R-wave to S1 and the time interval from S1 to S2 are relatively stable among individuals. This characteristic provides a reliable basis for the time alignment of multiple heart sound signals. For the first, second, and third sub-heart sound signals, the local signal-to-noise ratio (SNR) of each signal within the S1 and S2 windows is calculated. The calculation of the local SNR is based on the ratio of signal energy to background noise energy. A higher signal-to-noise ratio indicates better acquisition quality of the signal within the current heart sound cycle, and should be assigned a higher fusion weight.

[0040] Then, adaptive weighted fusion coefficients are constructed based on the local signal-to-noise ratio of each sub-heart sound signal. The weighted sub-heart sound signals are then superimposed point by point to obtain the fused heart sound signal.

[0041] Optionally, based on the sub-electrocardiogram signal, the first sub-heart sound signal, the second sub-heart sound signal, and the third sub-heart sound signal are fused to obtain a fused heart sound signal, including: Short-time Fourier transform is performed on the first sub-heart sound signal, the second sub-heart sound signal, and the third sub-heart sound signal to obtain the first complex spectrum of the first sub-heart sound signal, the second complex spectrum of the second sub-heart sound signal, and the third complex spectrum of the third sub-heart sound signal. For any sub-heart sound signal, the complex spectrum corresponding to the sub-heart sound signal is enhanced to obtain the enhanced complex spectrum; Based on the sub-ECG signal, the confidence gating weights of electrical events are determined. Based on the confidence gating weights of electrical events, all enhanced complex spectra are weighted and fused to obtain the fused heart sound signal.

[0042] In this embodiment, before performing short-time Fourier transform on the first, second, and third sub-heart sound signals, time delay compensation is performed on these signals. GCC-PHAT is used to estimate the relative time delay between the three heart sounds and then compensate for it to establish cross-lobe phase consistency. Specifically, a two-channel complex spectrum product is constructed in the frequency domain. To suppress the interference of phase whitening amplitude fluctuations on the correlation peaks, the following was obtained. The peak position is used as the time delay estimate. Then, compensation and alignment are performed. The calculation formula is as follows: in, The time-domain cross-correlation function is obtained after the inverse Fourier transform. This is the frequency domain representation of the parameter channel signal. This represents the frequency domain of the target channel signal. For target channel signal The complex conjugate, This is a numerically stable term. It should be noted that the parameter channel signal and the target channel signal are one of the following sub-heart sound signals: the first sub-heart sound signal, the second sub-heart sound signal, and the third sub-heart sound signal.

[0043] Based on the time-domain cross-correlation function Its peak position is used as the time delay estimate This is the optimal time delay estimate for the two signals.

[0044] After time delay compensation of the first, second, and third sub-heart sound signals, short-time Fourier transforms are performed on them to obtain the first complex spectrum of the first sub-heart sound signal, the second complex spectrum of the second sub-heart sound signal, and the third complex spectrum of the third sub-heart sound signal. The formula for the short-time Fourier transform of each sub-heart sound signal is the same, and the calculation formula is as follows: in, For the first Section, No. Sub-heart sound signal in frame index Frequency Index The complex spectral value below, Including the first heart sound signal, the second heart sound signal, and the third heart sound signal. For the first Section, No. The signal segment corresponding to the sampling point of the heart sound signal. For the preset time window function, The position of the current point within this frame, in order to Centered on the signal, local signals are weighted to suppress spectral leakage. These are complex exponential basis functions used to extract components of different frequencies. The step size.

[0045] It should be noted that, based on the corresponding frame length and step size, the set of sampling point indices for each segment and the truncated signal for each segment are determined, i.e., the sampling points included in each segment and the signals of the corresponding sampling points. The calculation formula is as follows: in, For the first The set of sampling point indices for the segment. This represents the signal at the sampling point corresponding to the m-th segment. The index number of the sampling point. For the first Segment number, Step size, The frame length. To unify the signal sequence on the sampling grid, i.e., the output sequence after resampling, For the indicator function, when the sampling point index Belongs to set When the condition is met, the value is 1; otherwise, it is 0. Use an indicator function. Put the corresponding signal sequence in the first Segment Index Set Samples within the set are retained, while samples outside the set are set to zero.

[0046] For any sub-heart sound signal, the corresponding complex spectrum is enhanced to obtain the enhanced complex spectrum. This enhancement can be achieved by constructing a gain function in the complex domain, which adaptively adjusts based on the amplitude and phase characteristics of the sub-heart sound signal's complex spectrum. For example, for frequency bands or specific frequency components in the complex spectrum that reflect heart sound characteristics, their amplitude gain is increased to enhance signal energy; for frequency bands with severe noise interference, the gain is appropriately reduced to suppress noise. In terms of phase processing, the original phase information can be preserved or fine-tuned to ensure that the enhanced complex spectrum accurately reflects the temporal characteristics of the heart sound signal in phase, providing a more reliable complex domain data foundation for subsequent feature extraction and analysis.

[0047] Based on the ECG signal, the confidence gating weights for electrical events are determined. These weights can be determined using existing techniques, and this embodiment does not impose any limitations. Based on these confidence gating weights, all enhanced complex spectra are weighted and fused to obtain a fused heart sound signal. First, the weight value for each enhanced complex spectrum is determined, which can be calculated based on the consistency coefficient and prior quality. The consistency coefficient is determined based on the self-power spectrum and cross-power spectrum. The self-power spectrum represents the energy of a single signal at a time-frequency point, and the cross-power spectrum represents the correlation between two signals at a time-frequency point. The formula for calculating the consistency coefficient is as follows: in, For the first Section, No. Sub-heart sound signal in frame index Frequency Index Consistency coefficient under the following conditions For the first Section, No. The heart sound signal and the first Sub-heart sound signal in frame index Frequency Index The cross-power spectrum below, For the first Section, No. Sub-heart sound signal in frame index Frequency Index The self-power spectrum below, For the first Section, No. Sub-heart sound signal in frame index Frequency Index The self-power spectrum below, To avoid parameters with denominators of 0 or unstable values, ε takes the same value in all formulas of this application, preferably. .

[0048] Prior quality is a signal quality assessment index, which can be calculated based on the confidence gating weights of the electrical events in the corresponding ECG signal segment. Then, the weight value is calculated based on the consistency coefficient and prior quality. The calculation formula is as follows: in, For the first Section, No. Sub-heart sound signal in frame index Frequency Index The weight values ​​below, For the first Section, No. The prior quality of the child heart sound signal, For the first Section, No. Sub-heart sound signal in frame index Frequency Index Consistency coefficient under the following conditions For the first Section, No. The prior quality of the child heart sound signal, For the first Section, No. Sub-heart sound signal in frame index Frequency Index Consistency coefficient under the following conditions For the first Section, No. The prior quality of the child heart sound signal, For the first Section, No. The integrated energy of the sub-heart sound signal For the first The segment corresponds to the electrocardiogram reference energy index of the electrocardiogram signal. For electro-acoustic coupling weighting coefficients, To prevent regular terms with a denominator of zero, ε, , , This is a hyperparameter.

[0049] The enhanced complex spectra are weighted and fused according to their weight values ​​to obtain the fused heart sound signal. The calculation formula for the fused heart sound signal is as follows: in, For the first Segment, in frame index Frequency Index The fused heart sound signal below, i.e. the first The fused heart sound signal after fusing the three heart sound signals. No. Section, No. Heart sound signal, No. Each scale in the frame index Frequency Index The weight values ​​below, For the first The scale weight values ​​for each scale. For the first Section, No. Heart sound signal, No. Each scale in the frame index Frequency Index The enhanced complex spectrum below, For the first Section, No. The scale weight values ​​for each scale. For the first Section, No. Quality assessment indicators at various scales For the first Section, No. Quality assessment indicators at various scales.

[0050] In this embodiment, the scale weight value is used to adaptively adjust the contribution of each scale signal in the fusion process according to the quality of each scale signal, so that the fused heart sound signal can retain the effective information in each scale to the maximum extent, thereby effectively improving the accuracy of the fused heart sound signal.

[0051] Optionally, the complex spectrum corresponding to the sub-heart sound signal is enhanced to obtain an enhanced complex spectrum, including: The signal power of the sub-heart tone signal is calculated based on the complex spectrum. Based on the signal power, an adaptive enhancement gain is constructed; Based on the adaptive enhancement gain, the complex spectrum of the sub-heart sound signal is enhanced to obtain the enhanced complex spectrum.

[0052] In this embodiment, the signal power of the sub-heart tone signal is calculated based on the complex spectrum, and the calculation formula is as follows: in, For the first Section, No. Sub-heart sound signal in frame index Frequency Index Signal power at time and frequency points, For the first Section, No. Sub-heart sound signal in frame index Frequency Index Complex spectrum values ​​at time-frequency points.

[0053] Based on the signal power, an adaptive enhancement gain is constructed, and the calculation formula is as follows: in, For the first Segment corresponding to sub-heart sound signal in frame index Frequency Index Adaptive enhancement gain at time and frequency points For the first Segment corresponding to sub-heart sound signal in frame index Frequency Index High gain coefficient at time and frequency points For the first Segment corresponding to sub-heart sound signal in frame index Frequency Index Signal power at time and frequency points, For the first Segment corresponding to sub-heart sound signal in frame index Frequency Index Low gain coefficient at time and frequency points.

[0054] Based on the adaptive enhancement gain, the complex spectrum of the sub-heart sound signal is enhanced to obtain the enhanced complex spectrum. The calculation formula is as follows: in, For the first Section, No. Sub-heart sound signal in frame index Frequency Index Enhanced complex spectrum at time and frequency points For the first Segment corresponding to sub-heart sound signal in frame index Frequency Index Adaptive enhancement gain at time and frequency points For the first Section, No. Sub-heart sound signal in frame index Frequency Index The complex spectral value below.

[0055] S104: Calculate the evidence abnormality measure of the heart sound signal based on the fused heart sound signal and the sub-ECG signal within the corresponding segment.

[0056] In step S104, the evidence anomaly metric is a quantitative representation of the time-frequency consistency between the fused heart sound signal and the sub-ECG signal within the corresponding segment.

[0057] In this embodiment, based on the fused heart sound signal and the corresponding sub-ECG signal within the segment, an evidence anomaly metric for the heart sound signal is calculated. Simultaneous short-time Fourier transforms can be performed on the fused heart sound signal and the corresponding sub-ECG signal within the segment to obtain the heart sound time spectrum and the ECG time spectrum. The cross-spectral density of the two time spectra at corresponding time-frequency points is then calculated. The cross-spectral density reflects the joint energy distribution of the heart sound signal and the ECG signal at a specific time-frequency point.

[0058] In this embodiment, the evidence abnormality measure of the heart sound signal is calculated based on the fused heart sound signal and the sub-electrocardiogram signal within the corresponding segment. By introducing the electrocardiogram signal as a physiological synchronization reference, the normal physiological components and abnormal pathological components in the heart sound signal can be effectively distinguished.

[0059] Optionally, measures of evidence anomalousness include heart sound-electrocardiogram consistency residuals, gating leakage rate, leakage spectrum index, drift degree, and consistency degradation. Based on the fused heart sound signal and the corresponding sub-ECG signal within the segment, an evidence measure of abnormality in the heart sound signal is calculated, including: Obtain the electrical event confidence gating weights corresponding to the sub-ECG signals. Based on the electrical event confidence gating weights and the fused heart sound signals, calculate the heart sound-ECG consistency residuals, gating leakage rate, leakage spectrum index, drift degree, and consistency degradation.

[0060] In this embodiment, electrical event confidence gating weights are obtained based on the sub-ECG signals within the corresponding segment. These weights characterize the reliability of electrophysiological events occurring in the ECG signals and their effective constraint strength on heart sound analysis. The electrical event anchor time and confidence level are extracted from the sub-ECG signals within the time window. The electrical event anchor time includes the R-peak time and the T-peak time. The electrical event anchor time is mapped to the intra-segment frame index, and the electrical event confidence gating weights are generated based on the intra-segment frame index corresponding to the electrical event anchor time.

[0061] Based on the electrical event confidence gating weights and the fused heart sound signals, the heart sound-ECG consistency residual, gating leakage rate, leakage spectral index, drift, and consistency degradation were calculated. The heart sound-ECG consistency residual characterizes the synchronicity between the heart sound and ECG signals, reflecting the degree of difference and synchronization deviation between them in the time domain, feature points, and waveform morphology. The gating leakage rate characterizes the extent to which effective heart sound signal energy leaks outside the preset time gate, reflecting the precise constraint capability of the gating constraint on the effective signal. The leakage spectral index distinguishes between heart sound spillover and non-cooperative perturbation during leakage. The drift characterizes the temporal stability of the energy distribution in the valve region. The consistency degradation characterizes the decrease in the reliability of heart sound signal consistency across valve regions and multiple channels under gating, reflecting the degradation of the physiological synchronicity of the heart sound signal.

[0062] When calculating the residuals of heart sound-ECG consistency, the acoustic energy of each frame within a segment is first defined, as shown in the following formula: in, For the first Segment fusion of heart sound signals, frame index The corresponding acoustic energy, No. Segment, in frame index Frequency Index The fusion of heart sounds below, This is the set of effective frequencies.

[0063] Based on acoustic energy, the distribution of acoustic events is calculated using the following formula: in, For the first Segment Fusion Heart Sound Signal Frame Index The distribution of acoustic events, For the first Segment fusion of heart sound signals, frame index The corresponding acoustic energy, No. Segment, in frame index Frequency Index The fusion of heart sounds below, For the effective frequency set, To prevent A parameter that is zero.

[0064] The distribution of electrical events is calculated using the following formula: in, For the first Joke ECG signal frame index The distribution of electrical events, For frame indexing The confidence gating weight of electrical events To prevent A parameter that is zero.

[0065] Based on the acoustic event distribution and the electrical event distribution, the residuals of heart sound-electrocardiogram consistency are calculated using the following formula: in, For the residuals of heart sound-electrocardiogram consistency, For the first Segment Fusion Heart Sound Signal Frame Index The distribution of acoustic events, For the first Joke ECG signal frame index The distribution of electrical events, To prevent A parameter that is zero. The smaller the value of the residual for heart sound-electrocardiogram consistency, the better the physiological synchronicity between heart sound and electrocardiogram.

[0066] When calculating the gate leakage rate, it can be calculated based on the ratio of external energy to internal energy. The calculation formula is as follows: in, For gate leakage rate, For gated external energy, For gated internal energy, For frame indexing The confidence gating weight of electrical events No. Segment, in frame index Frequency Index The fusion of heart sounds below, To prevent The parameter is zero. The smaller the value of the gating leakage rate, the more concentrated the effective signal energy is within the preset time gate, the more precise the gating constraint, and the purer the extracted physiological characteristics.

[0067] The formula for calculating the leakage spectrum index is as follows: in, The leakage spectrum index. For frame indexing The confidence gating weight of electrical events No. Segment, in frame index Frequency Index The fusion of heart sounds below, For gated external energy, To prevent A parameter that is zero. The smaller the value, the more it tends to be dominated by non-cooperative disturbances.

[0068] Under the gated energy constraint, the channel weights are compressed into lobe region contribution components, calculated as follows: in, For the first The contribution components of the segment and the c-th sub-heart sound signal to the corresponding valve region, For the first Section, No. Sub-heart sound signal in frame index Frequency Index The weight values ​​below, No. Segment, in frame index Frequency Index The fusion of heart sounds below, For frame indexing The confidence gating weight of electrical events.

[0069] Forming the lobe region contribution vector; .in, It contributes a component to the valve region corresponding to the first sub-heart sound signal. It contributes a component to the valve region corresponding to the second sub-heart sound signal. It contributes a component to the valve region corresponding to the third sub-heart sound signal.

[0070] The drift is calculated based on the lobe contribution vector, using the following formula: in, This is the smoothing coefficient for baseline recursion, with a value between 0 and 1. The larger the value, the smoother the baseline update and the stronger the resistance to transient noise; The smaller the value, the faster the baseline response and the more sensitive it is to short-term changes. The optimal value is 0.95, which strikes a balance between individualized stability and short-term adaptability. For the first The contribution vector of the lobe region after segment smoothing. Contribute vector to the lobe region. For the first The contribution vector of the lobe region after segment smoothing. This refers to the drift degree. The smaller the drift degree, the closer the energy distribution of the current segment lobe region is to the individual baseline, the more stable the temporal sequence, and the more consistent the physiological pattern.

[0071] The consistency degradation is calculated based on the consistency coefficient and the channel weight value, using the following formula: in, For the first The degree of uniformity degradation of segmental heart sound signals. For the first Section, No. Sub-heart sound signal in frame index Frequency Index The weight values ​​below, For the first Section, No. Sub-heart sound signal in frame index Frequency Index The complex spectral value below, To prevent parameters with a denominator of 0, the closer the consistency degradation is to 1, the higher the consistency of the heart sound signal across the transvalvular region and multiple channels, and the better the physiological synchronicity.

[0072] An anomaly metric for evidence is constructed based on the residuals of heart sound-electrocardiogram consistency, gating leakage rate, leakage spectrum index, drift, and consistency degradation. The formula for the anomaly metric is as follows: in, For the residuals of heart sound-electrocardiogram consistency, For gate leakage rate, The leakage spectrum index. For drift degree, This represents the consistency degradation amount.

[0073] In another embodiment, the evidence anomaly measurement may further include heart sound-ECG time alignment deviation, which characterizes the time delay between the fused heart sound signal and the sub-ECG signal within the corresponding segment, i.e., the time difference between the feature points of the fused heart sound signal and the feature points of the sub-ECG signal within the corresponding segment.

[0074] S105: Based on the evidence anomaly measure, perform risk prediction on the personnel to be tested and obtain the risk prediction results.

[0075] In step S105, the risk prediction result is the prediction result of the health risk of the person to be tested.

[0076] In this embodiment, risk prediction for the test subjects is performed based on the evidence anomaly metric. A pre-trained risk prediction model can be used for this prediction. During training, the model is trained using heart sound and electrocardiogram signals from multiple test subjects, as well as their risk labels. This embodiment does not impose limitations on this method. The evidence anomaly metric is input into the pre-trained risk prediction model, which outputs a risk prediction result, categorized as either healthy or unhealthy.

[0077] Optionally, risk prediction is performed on the individuals to be tested based on anomaly measures of evidence, resulting in risk prediction results, including: Perform a bounded mapping on each evidence anomaly measure to obtain the mapping value of each evidence anomaly measure; The mapping values ​​of each evidence anomaly metric are aggregated to obtain the risk value; Risk prediction is performed based on the risk value, and the risk prediction result is obtained.

[0078] In this embodiment, a bounded mapping is performed on each measure of evidence anomaly. Specifically, bounded mappings are performed on the heart sound-ECG consistency residual, gating leakage rate, leakage spectrum index, drift, and consistency degradation, respectively, to obtain the mapped values ​​for each measure of evidence anomaly: the mapped values ​​for the heart sound-ECG consistency residual, gating leakage rate, leakage spectrum index, drift, and consistency degradation. The calculation formulas are as follows: in, Let j be the mapping value of the evidence anomaly metric. For the j-th evidence anomaly measure, This is a parameter, and its value should be determined based on the actual situation.

[0079] The mapping values ​​of each evidence anomaly metric are aggregated to obtain the risk value. The formula for calculating the risk value is as follows: in, Let m be the risk value of segment m. Let j be the mapping value of the evidence anomaly metric. It is an aggregate function.

[0080] When making risk predictions based on risk values, when the risk value is in a continuous range... If all segments are not lower than the first preset threshold, the risk prediction result is determined to be unhealthy.

[0081] It should be noted that when the risk prediction result is determined to be unhealthy, an alarm can also be triggered. This alarm can be triggered when the risk value is continuously... When all segments are below the second preset threshold, the alarm state is exited to create hysteresis and prevent repeated alarms. The first preset threshold is greater than the second preset threshold.

[0082] In this embodiment, by measuring the anomalies of multi-source heterogeneous evidence and performing bounded mapping and aggregation of the anomaly measurement results of each piece of evidence, a quantitative assessment and dynamic prediction of system risk is achieved without relying on complex prior assumptions or manual intervention, thus ensuring the timeliness of risk warning.

[0083] See Figure 2 , Figure 2 This is a schematic diagram of a risk prediction device based on heart sound and electrocardiogram (ECG) according to an embodiment of this application. The risk prediction device and the risk prediction method based on heart sound and ECG correspond one-to-one. For ease of explanation, only the parts relevant to this embodiment are shown. The risk prediction device includes a main control module 400, three-lead electrode pads 100, and multiple electret microphones 300. The three-lead electrode pads 100 are connected to the main control module 400 via retractable lead wires 200. The retractable lead wires 200 are lead wires with retractable function, used to realize the electrical connection between the three-lead electrode pads and the main control module. For example, the retractable lead wires 200 can be elastic retractable wires, spiral retractable wires, or flexible retractable cables. Each electret microphone 300 is connected to the main control module 400 via a flexible retractable connection structure 310. The flexible retractable connection structure 310 is a connector with elastic retractable characteristics, used to realize the flexible connection between the electret microphone 300 and the main control module 400. For example, the flexible retractable connection structure 310 can be a flexible circuit board, elastic wire, spring wire, etc.

[0084] It should be noted that the main control module 400 includes a main control chip, an analog-to-digital converter chip electrically connected to the main control chip, a preamplifier connected to the analog-to-digital converter, and a data storage unit. The main control chip is used to execute the aforementioned risk prediction method. The analog-to-digital converter chip is used to convert the analog signal output from the three-lead electrode plate into a digital signal. The preamplifier is used to amplify the analog signal. The data storage unit is used to store the data generated during the execution of the risk prediction method.

[0085] The three-lead electrode pad 100 is used to collect the electrocardiogram (ECG) signal of the subject and transmit it to the main control module 400. The three-lead electrode pad 100 is a common circular electrode pad found in the market. The base of the three-lead electrode pad 100 is made of non-woven fabric with ventilation holes, providing a larger contact area while ensuring comfort and breathability during prolonged wear. The outer part of the three-lead electrode pad 100 uses a bio-gel to ensure stable contact with the skin surface, reducing contact impedance fluctuations and the risk of electrode detachment caused by skin laxity, sensitivity, or sweat, and minimizing artifacts that interfere with subsequent abnormal signal identification and risk warning judgment. The three-lead electrode pad 100 is connected to the analog-to-digital converter chip in the main control module 400 via a retractable lead wire 200, transmitting the collected ECG signal of the subject to the main control module 400.

[0086] Multiple electret microphones 300 are used to collect the first heart sound signal from the aortic valve auscultation area, the second heart sound signal from the mitral valve auscultation area, and the third heart sound signal from the tricuspid valve auscultation area, and transmit the first, second, and third heart sound signals to the main control module 400. This embodiment includes three electret microphones 300, placed in the aortic valve auscultation area, mitral valve auscultation area, and tricuspid valve auscultation area, respectively. This multi-point signal acquisition effectively improves the quality and accuracy of heart sound signal acquisition, making it more suitable for long-term home monitoring and intelligent diagnostic assistance for cardiovascular diseases such as heart failure. The electret microphones 300 are connected to the main control chip in the main control module 400 via a flexible telescopic connection structure 310 to ensure a stable fit and body surface adaptation during wear.

[0087] The main control module 400 is used to execute any of the above-mentioned risk prediction methods based on heart sounds and electrocardiograms.

[0088] See Figure 3 , Figure 3This is another structural schematic diagram of a risk prediction device based on heart sound and electrocardiogram (ECG) according to an embodiment of this application. The risk prediction device also includes an ECG lead wire telescopic module 500, which includes a telescopic handle 510 and a winding wheel 520. The telescopic handle 510 is used to drive the winding wheel 520 to rotate, and the winding wheel 520 is used to wind and unwind the retractable lead wire 200. The ECG lead wire telescopic module 500 is disposed on the housing of the main control module 400 and connected to the retractable lead wire 200. When the telescopic handle 510 rotates around the operating axis, it drives the winding wheel 520 to rotate in both directions, realizing the winding and unwinding of the retractable lead wire 200, and can slide axially between the unlocking position and the locking position. When in the unlocking position, the locking mechanism separates from the locking gear on the winding wheel, and the winding wheel 520 is in a freely rotating state, allowing continuous adjustment of the length of the retractable lead wire 200. When in the locking position, the locking mechanism engages with the locking gear, mechanically limiting the rotation angle of the winding wheel 520, thereby extending the adjusted retractable lead wire 200. This allows for individualized adjustment of the effective length of the retractable lead wire 200 and the electrode spacing for test subjects of different ages, body types, and genders, thereby reducing the incidence of wire sag, dragging, and tangling, and minimizing the risk of accidental electrode detachment due to mechanical pulling.

[0089] See Figure 4 , Figure 4 This is a schematic diagram of a risk prediction device based on heart sound and electrocardiogram according to an embodiment of this application, taken from different perspectives. Figure 4 (a), (b), (c), and (d) are schematic diagrams of the risk prediction device based on heart sound and electrocardiogram from different perspectives. They are used to clearly and completely show the overall structure of the device, the relative positional relationship of each component, the connection relationship, and the assembly relationship from multiple perspectives, so as to facilitate the understanding of the technical solution of the present invention.

[0090] It should be noted that the risk prediction device may also include a wireless transmission module, which synchronously transmits the collected electrocardiogram signals, heart sound signals and risk prediction results to the host computer in real time, supporting remote continuous monitoring, real-time early warning push and intelligent diagnostic auxiliary analysis.

[0091] See Figure 5 , Figure 5 This is a schematic diagram of wearing a risk prediction device based on heart sound and electrocardiogram provided in an embodiment of this application.

[0092] It should be noted that the information interaction and execution process between the above-mentioned units are based on the same concept as the method embodiments of this application. For details on their specific functions and technical effects, please refer to the method embodiments section, which will not be repeated here.

[0093] Figure 6This is a schematic diagram of the structure of a computer device provided in one embodiment of this application. For example... Figure 6 As shown, the computer device of this embodiment includes: at least one processor ( Figure 6 Only one is shown in the diagram), a memory, and a computer program stored in the memory and executable on at least one processor, wherein the processor executes the computer program to implement the steps in any of the above-described risk prediction method embodiments.

[0094] This computer device may include, but is not limited to, a processor and memory. Those skilled in the art will understand that... Figure 6 The examples of computer devices are merely examples and do not constitute a limitation on computer devices. Computer devices may include more or fewer components than shown, or combinations of certain components, or different components, such as network interfaces, displays, and input devices.

[0095] The processor referred to can be a CPU, but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor.

[0096] Memory includes readable storage media, internal memory, etc., wherein internal memory can be the RAM of a computer device, providing an environment for the operation of the operating system and computer-readable instructions stored in the readable storage media. The readable storage media can be the hard drive of a computer device, or in other embodiments, it can be an external storage device of the computer device, such as a plug-in hard drive, Smart Media Card (SMC), Secure Digital (SD) card, or Flash Card. Furthermore, memory can include both internal storage units and external storage devices of the computer device. Memory is used to store the operating system, applications, bootloader, data, and other programs, such as program code for computer programs. Memory can also be used to temporarily store data that has been output or will be output.

[0097] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above device can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here. If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of this application can be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the above method embodiments. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. A computer-readable medium can include at least: any entity or device capable of carrying computer program code, a recording medium, a computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media. Examples include USB flash drives, portable hard drives, magnetic disks, or optical disks. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electrical carrier signals or telecommunication signals.

[0098] The implementation of all or part of the processes in the methods of the above embodiments can also be accomplished by a computer program product. When the computer program product is run on a computer device, it causes the computer device to execute the steps in the above method embodiments.

[0099] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0100] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0101] In the embodiments provided in this application, it should be understood that the disclosed apparatus / computer devices and methods can be implemented in other ways. For example, the apparatus / computer device embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0102] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0103] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A risk prediction method based on heart sounds and electrocardiograms, characterized in that, The risk prediction method includes: Acquire the heart sound signals and electrocardiogram signals of the subject to be tested. The heart sound signals include the first heart sound signal in the aortic valve auscultation area, the second heart sound signal in the mitral valve auscultation area, and the third heart sound signal in the tricuspid valve auscultation area. According to a preset time window, the first channel heart sound signal, the second channel heart sound signal, the third channel heart sound signal and the electrocardiogram signal are segmented to obtain multiple segments of first sub-heart sound signals of the first channel heart sound signal, multiple segments of second sub-heart sound signals of the second channel heart sound signal, multiple segments of third sub-heart sound signals of the third channel heart sound signal and multiple segments of electrocardiogram signals of the electrocardiogram signal; For the first sub-heart sound signal, the second sub-heart sound signal, the third sub-heart sound signal and the sub-electrocardiogram signal in a corresponding segment within any time window, the first sub-heart sound signal, the second sub-heart sound signal and the third sub-heart sound signal are fused according to the sub-electrocardiogram signal to obtain a fused heart sound signal; The step of fusing the first sub-heart sound signal, the second sub-heart sound signal, and the third sub-heart sound signal based on the sub-ECG signal to obtain a fused heart sound signal includes: Perform a short-time Fourier transform on the first sub-heart sound signal, the second sub-heart sound signal, and the third sub-heart sound signal to obtain the first complex spectrum of the first sub-heart sound signal, the second complex spectrum of the second sub-heart sound signal, and the third complex spectrum of the third sub-heart sound signal; For any sub-heart sound signal, the complex spectrum corresponding to the sub-heart sound signal is enhanced to obtain the enhanced complex spectrum; Based on the sub-ECG signal, the confidence gating weight of the electrical event is determined. Based on the confidence gating weight of the electrical event, all enhanced complex spectra are weighted and fused to obtain the fused heart sound signal. Based on the fused heart sound signal and the sub-ECG signal within the corresponding segment, the evidence abnormality measure of the heart sound signal is calculated; Risk prediction is performed on the individuals to be tested based on the aforementioned evidence anomaly measurement, and the risk prediction results are obtained.

2. The risk prediction method as described in claim 1, characterized in that, The step of segmenting the first heart sound signal, the second heart sound signal, the third heart sound signal, and the electrocardiogram signal according to a preset time window to obtain multiple segments of first sub-heart sound signals of the first heart sound signal, multiple segments of second sub-heart sound signals of the second heart sound signal, multiple segments of third sub-heart sound signals of the third heart sound signal, and multiple segments of electrocardiogram signals of the electrocardiogram signal includes: The first heart sound signal, the second heart sound signal, the third heart sound signal, and the electrocardiogram signal are converted into signal sequences on a unified sampling grid to obtain the first resampled heart sound signal sequence of the first heart sound signal, the second resampled heart sound signal sequence of the second heart sound signal, the third resampled heart sound signal sequence of the third heart sound signal, and the resampled electrocardiogram signal sequence of the electrocardiogram signal. According to a preset time window, the first resampled heart sound signal sequence, the second resampled heart sound signal sequence, the third resampled heart sound signal sequence, and the resampled electrocardiogram signal sequence are segmented to obtain multiple segments of first sub-heart sound signals of the first heart sound signal, multiple segments of second sub-heart sound signals of the second heart sound signal, multiple segments of third sub-heart sound signals of the third heart sound signal, and multiple segments of sub-electrocardiogram signals of the electrocardiogram signal.

3. The risk prediction method as described in claim 1, characterized in that, The enhancement process for the complex spectrum corresponding to the sub-heart sound signal to obtain the enhanced complex spectrum includes: The signal power of the sub-heart sound signal is calculated based on the complex spectrum. Based on the signal power, an adaptive enhancement gain is constructed; Based on the adaptive enhancement gain, the complex spectrum of the sub-heart sound signal is enhanced to obtain the enhanced complex spectrum.

4. The risk prediction method as described in claim 1, characterized in that, The measures of evidence anomalies include heart sound-electrocardiogram consistency residuals, gating leakage rate, leakage spectrum index, drift degree, and consistency degradation. The step of calculating the evidence anomalous measure of the heart sound signal based on the fused heart sound signal and the sub-ECG signal within the corresponding segment includes: Obtain the electrical event confidence gating weights corresponding to the sub-ECG signals, and calculate the heart sound ECG consistency residuals, gating leakage rate, leakage spectrum index, drift degree, and consistency degradation amount based on the electrical event confidence gating weights and the fused heart sound signals.

5. The risk prediction method as described in claim 4, characterized in that, The step of performing risk prediction on the person to be tested based on the evidence anomaly measurement to obtain the risk prediction result includes: Perform a bounded mapping on each evidence anomaly measure to obtain the mapping value of each evidence anomaly measure; The mapping values ​​of each evidence anomaly metric are aggregated to obtain the risk value; Risk prediction is performed based on the risk value to obtain the risk prediction result.

6. A risk prediction device based on heart sound and electrocardiogram, characterized in that, The risk prediction device includes a main control module, a three-lead electrode sheet, and multiple electret microphones. The three-lead electrode sheet is connected to the main control module via a retractable lead wire, and each electret microphone is connected to the main control module via a flexible retractable connection structure. The three-lead electrode pads are used to collect the electrocardiogram (ECG) signals of the person being tested and transmit the ECG signals to the main control module; Multiple electret microphones are used to collect the first heart sound signal from the aortic valve auscultation area, the second heart sound signal from the mitral valve auscultation area, and the third heart sound signal from the tricuspid valve auscultation area, and transmit the first heart sound signal, the second heart sound signal, and the third heart sound signal to the main control module; The main control module is used to execute the risk prediction method based on heart sound and electrocardiogram as described in any one of claims 1 to 5.

7. The risk prediction device as described in claim 6, characterized in that, The risk prediction device also includes an ECG lead retractable module, which includes a telescopic handle and a winding reel. The telescopic handle is used to drive the winding reel to rotate, and the winding reel is used to wind up and unwind the retractable lead.

8. A computer device, characterized in that, The computer device includes a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the risk prediction method as described in any one of claims 1 to 5.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the risk prediction method as described in any one of claims 1 to 5.

Images