Blood pressure prediction method and device and electronic equipment
By combining photovoltaic pulse wave gramography, electrocardiogram and heart sound signal, the time difference between key characteristic points in each signal combination is determined, which solves the problems of complex operation and inaccurate prediction of existing blood pressure monitoring methods, and achieves high-accurate blood pressure prediction.
Patent Information
- Application Number
- CN202510465193.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-04-14
- Publication Date
- 2025-06-27
AI Technical Summary
The existing blood pressure monitoring methods have problems such as complex operation, inability to achieve continuous real-time measurement, and high invasiveness. The photoplethysmography only uses pulse wave conduction time as a characteristic parameter to affect the accuracy of blood pressure prediction.
By obtaining the photovoltaic pulse wave gram signal, electrocardiogram signal and heart sound signal, the key feature point moments in each cardiac cycle are determined, and the time difference between the key feature points in each signal combination is calculated to predict the current blood pressure value.
This method can fully reflect blood pressure values, reduce the limitations of a single signal, improve the accuracy of blood pressure value prediction, and achieve non-invasive, real-time continuous blood pressure monitoring.
Smart Images
Figure CN120203544A_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to blood pressure monitoring technology, and more particularly, to a blood pressure prediction method, apparatus, and electronic device. Background Art
[0002] Blood pressure is a measurement of the force exerted by blood on the vessel wall per unit area and is one of the important indicators of human physiological health. The change characteristics of blood pressure reflect the state of the human cardiovascular system. Therefore, blood pressure monitoring is crucial for the diagnosis and treatment of many cardiovascular diseases.
[0003] Common blood pressure measurement methods include auscultation method, oscillometric method, direct intra-arterial measurement method, photoplethysmography method, etc. The mercury sphygmomanometer is representative of the auscultation method. This method compresses the artery with an inflatable cuff to block blood flow and completes a complete measurement by listening to the Korotkoff sounds generated when blood flow resumes during cuff deflation. Although this method has high accuracy, it is complex to operate, greatly affected by the operator's subjective judgment, and cannot achieve continuous real-time measurement. The arm-type electronic sphygmomanometer is based on the oscillometric method and also needs to use the cuff inflation to block arterial blood flow. Then, during deflation, the gas pressure in the cuff is detected and the weak pulse wave is extracted, and then the blood pressure is estimated according to the relationship between the pulse wave amplitude and the cuff pressure. The direct intra-arterial measurement method directly measures blood pressure by inserting an arterial catheter. Although the measurement result is accurate, it is only applicable to intensive care. This invasive operation has risks of infection and thrombosis and is not suitable for daily use.
[0004] The physiological signal analysis method represented by photoplethysmography has the advantages of non-invasive, portable, and real-time continuous monitoring. This method extracts characteristic parameters and analyzes the correlation between blood pressure values and characteristic parameters, and then predicts the blood pressure value. Currently, for this method, the extracted characteristic parameter is the pulse wave transit time PWTT (Pulse Wave Transit Time), and the characteristic parameter is relatively single. When predicting blood pressure only using the PWTT parameter, it affects the accuracy of blood pressure prediction. Summary of the Invention
[0005] An object of the present invention is to provide a new technical solution for a blood pressure prediction method.
[0006] According to a first aspect of the present invention, there is provided a blood pressure prediction method, including:
[0007] Obtaining a photoplethysmogram signal, an electrocardiogram signal, and a heart sound signal within a preset time length;
[0008] Based on each cardiac cycle within the preset time length, determining the time when key feature points are located in the photoplethysmogram signal, the time when key feature points are located in the electrocardiogram signal, and the time when key feature points are located in the heart sound signal;
[0009] Based on the moments of key feature points in the photoplethysmogram (PPG) signals of each cardiac cycle, the moments of key feature points in the electrocardiogram (ECG) signals, and the moments of key feature points in the heart sound signals, determine the time differences between the key feature points in all pairwise combinations of signals within each cardiac cycle;
[0010] Predict the current blood pressure value based on the time differences between the key feature points in all pairwise combinations of signals within each cardiac cycle.
[0011] Optionally, the key feature points in the PPG signals are the systolic peak points and valley points, the key feature points in the ECG signals are the peak points, and the key feature points in the heart sound signals are the first heart sound peak point and the second heart sound peak point.
[0012] Optionally, based on each cardiac cycle, the time differences between the key feature points in the PPG signals and the key feature points in the ECG signals are the time differences between the peak point and the systolic peak point, and the time difference between the peak point and the valley point.
[0013] The time differences between the key feature points in the PPG signals and the key feature points in the heart sound signals are the time differences between the first heart sound peak point and the systolic peak point, the time difference between the first heart sound peak point and the valley point, the time difference between the second heart sound peak point and the systolic peak point, and the time difference between the second heart sound peak point and the valley point.
[0014] The time differences between the key feature points in the ECG signals and the key feature points in the heart sound signals are the time differences between the peak point and the first heart sound peak point, and the time difference between the peak point and the second heart sound peak point.
[0015] Optionally, the predicting the current blood pressure value based on the time differences between the key feature points in all pairwise combinations of signals within each cardiac cycle includes:
[0016] Determine the statistical characteristic values of the time differences between the key feature points in all pairwise combinations of signals within each cardiac cycle, where the statistical characteristic values are at least two of the mean, standard deviation, maximum value, minimum value, median, kurtosis, skewness, and coefficient of variation;
[0017] Predict the current blood pressure value based on the statistical characteristic values of the time differences between the key feature points in all pairwise combinations of signals and the corresponding regression model fitting coefficients.
[0018] Optionally, the method further includes:
[0019] Obtain the statistical eigenvalue of the time difference between key feature points in all pairwise combination signals corresponding to historical cardiac cycles and the actual blood pressure value corresponding to the historical cardiac cycle;
[0020] Based on the linear regression model, determine the fitting coefficients of each regression model according to the statistical eigenvalue of the time difference between key feature points in all pairwise combination signals corresponding to the historical cardiac cycle and the actual blood pressure value corresponding to the historical cardiac cycle.
[0021] Optionally, the method further includes:
[0022] Determine the correlation coefficient between each statistical eigenvalue of the time difference between key feature points in all pairwise combination signals and the blood pressure value;
[0023] Based on all pairwise combination signals, obtain the statistical eigenvalues corresponding to the correlation coefficients exceeding a preset threshold; wherein,
[0024] The predicting the current blood pressure value according to the statistical eigenvalue of the time difference between key feature points in all pairwise combination signals and the corresponding regression model fitting coefficient includes:
[0025] Predict the current blood pressure value according to the statistical eigenvalues corresponding to the correlation coefficients exceeding the preset threshold in all pairwise combination signals and the corresponding regression model fitting coefficients.
[0026] Optionally, before determining the time points where key feature points are located in the photoplethysmogram signal, the electrocardiogram signal, and the heart sound signal based on each cardiac cycle within the preset time length, the method further includes:
[0027] Perform normalization processing and filtering processing on the photoplethysmogram signal, the electrocardiogram signal, and the heart sound signal within the preset time length respectively to obtain the processed photoplethysmogram signal, electrocardiogram signal, and heart sound signal.
[0028] According to the second aspect of the present invention, there is provided a blood pressure prediction device, including:
[0029] An acquisition module, configured to acquire a photoplethysmogram signal, an electrocardiogram signal, and a heart sound signal within a preset time length;
[0030] A key feature point determination module, configured to determine the time points where key feature points are located in the photoplethysmogram signal, the electrocardiogram signal, and the heart sound signal based on each cardiac cycle within the preset time length;
[0031] A time difference determination module, configured to determine the time differences between key feature points in all pairwise combinations of signals within each cardiac cycle according to the time points of the key feature points in the photoplethysmogram signals of each cardiac cycle, the time points of the key feature points in the electrocardiogram signals, and the time points of the key feature points in the heart sound signals;
[0032] A blood pressure value prediction module, which predicts the current blood pressure value according to the time differences between the key feature points in all pairwise combinations of signals within each cardiac cycle.
[0033] According to a third aspect of the present invention, there is provided a blood pressure prediction device, including a memory and a processor, where the memory stores a computer program, and the computer program is used to control the processor to operate to execute the method according to any one of the first aspects.
[0034] According to a fourth aspect of the present invention, there is provided an electronic device, including the blood pressure prediction device according to the second aspect or the third aspect, a sensor for collecting photoplethysmogram signals, a sensor for collecting electrocardiogram signals, and a sensor for collecting heart sound signals, where
[0035] The sensor for collecting photoplethysmogram signals, the sensor for collecting electrocardiogram signals, and the sensor for collecting heart sound signals are all connected to the blood pressure prediction device.
[0036] The blood pressure prediction method provided by the present invention determines the time differences between key feature points from pairwise combination signals of the three signals, namely the photoplethysmogram signal, the electrocardiogram signal, and the heart sound signal, can comprehensively reflect the blood pressure value, reduce the limitations of a single signal, and improve the accuracy of blood pressure value prediction.
[0037] Through the following detailed description of the exemplary embodiments of the present specification with reference to the accompanying drawings, the features and advantages of the embodiments of the present specification will become clear. Description of the Drawings
[0038] The drawings incorporated in the specification and constituting a part of the specification illustrate the embodiments of the present specification, and together with the description are used to explain the principles of the embodiments of the present specification.
[0039] Figure 1 is a schematic flowchart of a blood pressure prediction method according to an embodiment of the present invention.
[0040] Figure 2 is a schematic diagram of three different types of signals according to an embodiment of the present invention.
[0041] Figure 3It is a schematic flow chart of obtaining blood pressure values based on predictions of three different types of signals according to an embodiment of the present invention.
[0042] Figure 4 It is a schematic structural diagram of a blood pressure prediction device according to an embodiment of the present invention.
[0043] Figure 5 It is a schematic structural diagram of a blood pressure prediction device according to an embodiment of the present invention.
[0044] Figure 6 It is a schematic structural diagram of an electronic device according to an embodiment of the present invention. Detailed implementation manners
[0045] Now, various exemplary embodiments of the present specification will be described in detail with reference to the accompanying drawings.
[0046] The following description of at least one exemplary embodiment is merely illustrative in nature and in no way serves as a limitation to the embodiments of the present specification, their applications, or uses.
[0047] It should be noted that: Similar reference numerals and letters denote similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further discussed in subsequent drawings.
[0048] In an embodiment of the present invention, a blood pressure prediction method is provided. According to Figure 1 As shown, the blood pressure prediction method of this embodiment includes the following steps S110 to S140.
[0049] Step S110, obtain photoplethysmography signals, electrocardiogram signals, and phonocardiogram signals within a preset time length.
[0050] The preset time length can be set according to user requirements. The preset time length includes at least two complete cardiac cycles. A cardiac cycle refers to the time for the heart to complete one complete pulsation process, including two stages of cardiac contraction and relaxation.
[0051] Photoplethysmography signals (PPG), electrocardiogram signals (ECG), and phonocardiogram signals (PCG) are respectively collected by corresponding sensors.
[0052] Step S120, based on each cardiac cycle within the preset time length, determine the moments when key feature points are located in the photoplethysmography signal, the moments when key feature points are located in the electrocardiogram signal, and the moments when key feature points are located in the phonocardiogram signal.
[0053] The key feature points in the photoplethysmogram (PPG) signal are the peak systolic peak point and the valley point. The key feature points in the electrocardiogram (ECG) signal are the peak points. The key feature points in the phonocardiogram (PCG) signal are the first heart sound peak point and the second heart sound peak point. The key feature points in each type of signal have clear meanings in the medical field, are easy to detect and extract, can reduce the result deviation caused by noise interference, and provide an accurate data basis for subsequent blood pressure prediction.
[0054] Figure 2 is a schematic diagram of the photoplethysmogram signal, electrocardiogram signal, and phonocardiogram signal according to an embodiment of the present invention. According to Figure 2 As shown, based on each cardiac cycle, the key feature points in PPG include a peak systolic peak point (Peak) and a valley point (Onset), the key feature points in ECG include a peak point (R), and the key feature points in PCG include a first heart sound peak point (S1) and a second heart sound peak point (S2).
[0055] Step S130, according to the time points where the key feature points are located in the photoplethysmogram signal, electrocardiogram signal, and phonocardiogram signal of each cardiac cycle, determine the time differences between the key feature points in all pairwise combinations of signals within each cardiac cycle.
[0056] Based on each cardiac cycle, the time differences between the key feature points in the photoplethysmogram signal and the key feature points in the electrocardiogram signal are the time difference between the peak point and the peak systolic peak point, and the time difference between the peak point and the valley point. Combining Figure 2 , the time difference between the peak point and the peak systolic peak point is denoted as T R to Peak , the time difference between the peak point and the valley point is denoted as T R to Onset .
[0057] Based on each cardiac cycle, the time differences between the key feature points in the photoplethysmogram signal and the key feature points in the phonocardiogram signal are the time difference between the first heart sound peak point and the peak systolic peak point, the time difference between the first heart sound peak point and the valley point, the time difference between the second heart sound peak point and the peak systolic peak point, and the time difference between the second heart sound peak point and the valley point. Combining Figure 2 , the time difference between the first heart sound peak point and the peak systolic peak point is denoted as T S1 to Peak , the time difference between the first heart sound peak point and the valley point is denoted as T S1 toOnset The time difference between the peak point of the second heart sound and the peak point of the systolic peak is denoted as T S2 to Peakt The time difference between the peak point of the second heart sound and the trough point is denoted as T S2 to Onset 。
[0058] Based on each cardiac cycle, the time differences between the key feature points in the electrocardiogram signal and the key feature points in the heart sound signal are the time differences between the peak point and the peak point of the first heart sound, and the time difference between the peak point and the peak point of the second heart sound. Combining Figure 2 The time difference between the peak point and the peak point of the first heart sound is denoted as T R to S1 The time difference between the peak point and the peak point of the second heart sound is denoted as T R to S2 。
[0059] Step S140, predict the current blood pressure value according to the time differences between the key feature points in all pairwise combinations of signals within each cardiac cycle.
[0060] In some embodiments, step S140 specifically includes steps S141 to S142.
[0061] Step S141, determine the statistical characteristic values of the time differences between the key feature points in all pairwise combinations of signals within each cardiac cycle, where the statistical characteristic values are at least two of the mean, standard deviation, maximum value, minimum value, median, kurtosis, skewness, and coefficient of variation. This can extract useful information related to the blood pressure value from multiple dimensions from the time differences between the key feature points in all pairwise combinations of signals, improving the accuracy of blood pressure value prediction.
[0062] Based on each cardiac cycle, a total of 8 types of time differences are obtained, namely T R to Peak 、T R to Onset 、T S1 to Peak 、T S1 to Onset 、T S2 to Peakt 、T S2 to Onset, T R to S1 , T R to S2 . Within a preset time length, multiple values are obtained based on each of the above types of time differences. Based on each of the above types of time differences, statistical characteristic values are determined.
[0063] Step S142, based on the statistical characteristic values of the time differences between key feature points in all pairwise combinations of signals and the corresponding regression model fitting coefficients, predict the current blood pressure value.
[0064] By calculating the blood pressure value through the statistical characteristic values of the time differences between key feature points in all pairwise combinations of signals and the corresponding regression model fitting coefficients, non-invasive, non-destructive, real-time and continuous blood pressure monitoring can be achieved, which is convenient for users' daily use and improves the user experience.
[0065] Specifically, based on the following calculation formula, predict the current blood pressure value,
[0066]
[0067] where BP pre is the predicted current blood pressure value, F i is a certain statistical characteristic value of the time differences between key feature points in all pairwise combinations of signals, n is the number of all statistical characteristic values, S i is the regression model fitting coefficient corresponding to a certain statistical characteristic value of the time differences between key feature points in all pairwise combinations of signals, and ε is the error term.
[0068] When the statistical characteristic values include mean, standard deviation, maximum value, minimum value, median, kurtosis, skewness, and coefficient of variation, combined with the above calculation formula, F i includes the mean, standard deviation, maximum value, minimum value, median, kurtosis, skewness, and coefficient of variation corresponding to T R to Peak , the mean, standard deviation, maximum value, minimum value, median, kurtosis, skewness, and coefficient of variation corresponding to T R to Onset , the mean, standard deviation, maximum value, minimum value, median, kurtosis, skewness, and coefficient of variation corresponding to T S1 to Peak , the mean, standard deviation, maximum value, minimum value, median, kurtosis, skewness, and coefficient of variation corresponding to T S1 to Onset , the mean, standard deviation, maximum value, minimum value, median, kurtosis, skewness, and coefficient of variation corresponding to T S2 toPeakt The corresponding mean, standard deviation, maximum value, minimum value, median, kurtosis, skewness, coefficient of variation, T S2 to Onset The corresponding mean, standard deviation, maximum value, minimum value, median, kurtosis, skewness, coefficient of variation, T R to S1 The corresponding mean, standard deviation, maximum value, minimum value, median, kurtosis, skewness, coefficient of variation, T R to S2 The corresponding mean, standard deviation, maximum value, minimum value, median, kurtosis, skewness, coefficient of variation. The regression model fitting coefficient corresponding to each of the above eigenvalue is a pre-stored value and can be directly obtained.
[0069] The blood pressure prediction method provided by the embodiment of the present invention, through the photoplethysmogram signal, electrocardiogram signal and heart sound signal, determines the time difference between key feature points from the pairwise combined signals of these three signals, can comprehensively reflect the blood pressure value, reduce the limitation of a single signal, and improve the accuracy of blood pressure value prediction.
[0070] For the processing flow of the blood pressure prediction method shown in steps S110 - S140, it can be understood in combination with Figure 3 the schematic diagram shown.
[0071] According to Figure 3 shown, through the photoplethysmogram signal acquisition module, electrocardiogram signal acquisition module and heart sound signal acquisition module, the PPG signal, ECG signal and PCG signal are respectively acquired. The photoplethysmogram signal acquisition module, electrocardiogram signal acquisition module and heart sound signal acquisition module can be corresponding sensors.
[0072] For the PPG signal, based on each cardiac cycle, determine the time when the systolic peak point (Peak) is located and the time when the valley point (Onset) is located. For the ECG signal, based on each cardiac cycle, determine the time when the peak point (R) is located. For the PCG signal, based on each cardiac cycle, determine the time when the first heart sound peak point (S1) is located and the time when the second heart sound peak point (S2) is located.
[0073] Based on each cardiac cycle, determine the time difference T between the peak point and the systolic peak point R to Peak and the time difference T between the peak point and the valley point R to Onset . The T corresponding to each cardiac cycle R toPeak and T R to Onset , to form a PAT sequence.
[0074] Based on each cardiac cycle, determine the time difference T between the peak point of the first heart sound and the peak point of the systolic peak S1 to Peak and the time difference T between the peak point of the first heart sound and the trough point S1 to Onset and the time difference T between the peak point of the second heart sound and the peak point of the systolic peak S2 to Peakt and the time difference T between the peak point of the second heart sound and the trough point is denoted as T S2 to Onset . The T corresponding to each cardiac cycle S1 to Peak and T S1 to Onset and T S2to Peakt and T S2 to Onset , to form a PTT sequence.
[0075] Based on each cardiac cycle, determine the time difference T between the peak point of the wave and the peak point of the first heart sound R to S1 and the time difference T between the peak point of the wave and the peak point of the second heart sound R to S2 . The T corresponding to each cardiac cycle R to S1 and T R to S2 , to form a PEP sequence.
[0076] The PAT sequence includes two types of time differences, the PTT sequence includes four types of time differences, and the PEP sequence includes two types of time differences. Based on each type of time difference, determine the corresponding statistical characteristic values. According to the statistical characteristic values corresponding to each type of time difference and the corresponding regression model fitting coefficients, predict the current blood pressure value.
[0077] In some embodiments, the regression model fitting coefficients corresponding to each eigenvalue can be determined in the following manner: Obtain the statistical eigenvalues of the time differences between key feature points in all pairwise combinations of signals corresponding to historical cardiac cycles and the actual blood pressure values corresponding to the historical cardiac cycles; Based on a linear regression model, determine each regression model fitting coefficient according to the statistical eigenvalues of the time differences between key feature points in all pairwise combinations of signals corresponding to the historical cardiac cycles and the actual blood pressure values corresponding to the historical cardiac cycles. It should be noted that multiple groups of historical cardiac cycles are used to determine each regression model fitting coefficient, and each group of historical cardiac cycles includes multiple consecutive historical cardiac cycles.
[0078] Specifically, based on the system of equations composed of the following calculation formulas, calculate each regression model fitting coefficient.
[0079]
[0080] Among them, BP act is the actual blood pressure value, F i is the statistical eigenvalue of the time difference between key feature points in all pairwise combinations of signals corresponding to a certain group of historical cardiac cycles, S i is the regression model fitting coefficient to be obtained, and ε is the error term.
[0081] In some embodiments, the method further includes: determining the correlation coefficient between the statistical eigenvalue of the time difference between key feature points in all pairwise combinations of signals and the blood pressure value; Based on all pairwise combinations of signals, obtain the statistical eigenvalues corresponding to the correlation coefficients exceeding a preset threshold.
[0082] For example, based on the above three types of signals PPG, ECG, and PCG, a total of 8 types of time differences are obtained, namely T R to Peak 、T R to Onset 、T S1 to Peak 、T S1 to Onset 、T S2 to Peakt 、T S2 to Onset 、T R to S1 、T R to S2. Based on each type of time difference, the corresponding mean, standard deviation, maximum value, minimum value, median value, kurtosis, skewness, and coefficient of variation are determined. Based on each statistical characteristic value in each type of time difference, the correlation coefficient with the blood pressure value is determined. The larger the correlation coefficient, the greater the correlation between the corresponding statistical characteristic value in this type of time difference and the blood pressure value. The statistical characteristic values with greater correlation with the blood pressure value are retained, and the statistical characteristic values with less correlation with the blood pressure value are eliminated to improve the accuracy of blood pressure prediction.
[0083] The correlation coefficient between each statistical characteristic value in each type of time difference and the blood pressure value may be a Pearson correlation coefficient or other correlation coefficients.
[0084] In this embodiment, step S140 specifically includes: predicting the current blood pressure value according to the statistical characteristic value corresponding to the correlation coefficient of all the two-combination signals exceeding the preset threshold and the corresponding regression model fitting coefficient.
[0085] T R to Peak Take this type of time difference as an example and determine the corresponding mean, standard deviation, maximum value, minimum value, median, kurtosis, skewness, and coefficient of variation. R to Peak The correlation between the corresponding mean, standard deviation, maximum value, minimum value and blood pressure value exceeds the preset threshold, T R to Peak If the correlation between the corresponding median, kurtosis, skewness, coefficient of variation and blood pressure value does not exceed the preset threshold, then based on T R to Peak For this type of time difference, the statistical characteristic values involved in predicting the current blood pressure value are the corresponding mean, standard deviation, maximum value, and minimum value. The same is true for other types of time differences, so I will not go into details.
[0086] In some embodiments, before step S120 is executed, the method further includes: normalizing and filtering the photoelectric volumetric pulse wave signal, electrocardiogram signal and heart sound signal within a preset time length, respectively, to obtain the processed photoelectric volumetric pulse wave signal, electrocardiogram signal and heart sound signal. Specifically, Butterworth filters with different filtering ranges can be used to filter out noise and artifacts that affect the signal quality, thereby improving the overall quality of the signal. Butterworth filters include but are not limited to high-pass filters, low-pass filters and band-pass filters.
[0087] One embodiment of the present invention provides a blood pressure prediction device. Figure 4As shown, the blood pressure prediction device 400 includes an acquisition module 410, a key feature point determination module 420, a time difference determination module 430, and a blood pressure value prediction module 440.
[0088] The acquisition module 410 is configured to acquire photoplethysmogram signals, electrocardiogram signals, and heart sound signals within a preset time length.
[0089] The key feature point determination module 420 is configured to determine the moments of key feature points in the photoplethysmogram signal, the moments of key feature points in the electrocardiogram signal, and the moments of key feature points in the heart sound signal based on each cardiac cycle within the preset time length.
[0090] The time difference determination module 430 is configured to determine the time differences between key feature points in all pairwise combinations of signals within each cardiac cycle according to the moments of key feature points in the photoplethysmogram signal, the moments of key feature points in the electrocardiogram signal, and the moments of key feature points in the heart sound signal within each cardiac cycle.
[0091] The blood pressure value prediction module 440 is configured to predict the current blood pressure value according to the time differences between key feature points in all pairwise combinations of signals within each cardiac cycle.
[0092] In some embodiments, the key feature points in the photoplethysmogram signal are the systolic peak value points and the valley value points, the key feature points in the electrocardiogram signal are the peak value points, and the key feature points in the heart sound signal are the first heart sound peak value points and the second heart sound peak value points.
[0093] In some embodiments, based on each cardiac cycle, the time differences between the key feature points in the photoplethysmogram signal and the key feature points in the electrocardiogram signal are the time differences between the peak value point and the systolic peak value point, and the time differences between the peak value point and the valley value point.
[0094] The time differences between the key feature points in the photoplethysmogram signal and the key feature points in the heart sound signal are the time differences between the first heart sound peak value point and the systolic peak value point, the time differences between the first heart sound peak value point and the valley value point, the time differences between the second heart sound peak value point and the systolic peak value point, and the time differences between the second heart sound peak value point and the valley value point.
[0095] The time differences between the key feature points in the electrocardiogram signal and the key feature points in the heart sound signal are the time differences between the peak value point and the first heart sound peak value point, and the time differences between the peak value point and the second heart sound peak value point.
[0096] In some embodiments, the blood pressure value prediction module 440 is configured to determine a statistical feature value of the time differences between key feature points in all pairwise combinations of signals within each cardiac cycle, where the statistical feature value is at least two of mean, standard deviation, maximum value, minimum value, median, kurtosis, skewness, and coefficient of variation; and predict the current blood pressure value based on the statistical feature value of the time differences between key feature points in all pairwise combinations of signals and the corresponding regression model fitting coefficients.
[0097] In some embodiments, the device further includes a regression model fitting coefficient determination module.
[0098] The regression model fitting coefficient determination module is configured to obtain the statistical feature value of the time differences between key feature points in all pairwise combinations of signals corresponding to historical cardiac cycles and the actual blood pressure values corresponding to historical cardiac cycles; and determine each regression model fitting coefficient based on the linear regression model according to the statistical feature value of the time differences between key feature points in all pairwise combinations of signals corresponding to historical cardiac cycles and the actual blood pressure values corresponding to historical cardiac cycles.
[0099] In some embodiments, the device further includes a screening module.
[0100] The screening module is configured to determine the correlation coefficient between each statistical feature value of the time differences between key feature points in all pairwise combinations of signals and the blood pressure value; and obtain the statistical feature values corresponding to the correlation coefficients exceeding a preset threshold based on all pairwise combinations of signals. The blood pressure value prediction module 440 is configured to predict the current blood pressure value based on the statistical feature values corresponding to the correlation coefficients exceeding the preset threshold in all pairwise combinations of signals and the corresponding regression model fitting coefficients.
[0101] In some embodiments, the device further includes a preprocessing module. The preprocessing module is configured to perform normalization processing and filtering processing on the photoplethysmogram signal, electrocardiogram signal, and heart sound signal within a preset time length respectively, to obtain the processed photoplethysmogram signal, electrocardiogram signal, and heart sound signal.
[0102] An embodiment of the present invention provides a blood pressure prediction device. According to Figure 5 As shown, the blood pressure prediction device includes a memory 520 and a processor 510. The memory 520 stores a computer program, and the computer program is used to control the processor 510 to operate to execute the blood pressure prediction method provided according to any of the above embodiments.
[0103] An embodiment of the present invention provides an electronic device. According to Figure 6As shown, the electronic device includes the blood pressure prediction device provided in any of the above embodiments, a sensor for collecting photoplethysmography signals, a sensor for collecting electrocardiogram signals, and a sensor for collecting heart sound signals.
[0104] According to Figure 6 As shown, the sensor for collecting photoplethysmography signals, the sensor for collecting electrocardiogram signals, and the sensor for collecting heart sound signals are all connected to the blood pressure prediction device.
[0105] The electronic device can be any of the following: a watch, a bracelet, or a ring.
[0106] Each embodiment in this specification is described in a progressive manner. For the same or similar parts among the embodiments, reference can be made to each other. Each embodiment focuses on the differences from other embodiments.
[0107] The above describes specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be performed in a different order than in the embodiments and still achieve the desired result. Additionally, the processes depicted in the drawings do not necessarily require the specific order or sequential order shown to achieve the desired result. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0108] The embodiments of this specification can be systems, methods, and / or computer program products. The computer program products can include computer-readable storage media having computer instructions thereon for causing a processor to implement various aspects of the embodiments of this specification.
[0109] A computer-readable storage medium can be a tangible device that can retain and store computer instructions for use by a computer instruction execution device. A computer-readable storage medium can be, for example, but is not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer-readable storage medium include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanically encoded device, such as a punched card or raised structures in grooves storing computer instructions thereon, and any suitable combination of the foregoing. The computer-readable storage medium as used herein is not construed as an instantaneous signal per se, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagated through a waveguide or other transmission medium (e.g., an optical pulse through an optical fiber cable), or an electrical signal transmitted through a wire.
[0110] The computer instructions described herein can be downloaded from the computer-readable storage medium to various computing / processing devices, or downloaded to an external computer or external storage device through a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network can include a copper transmission cable, an optical fiber transmission, a wireless transmission, a router, a firewall, a switch, a gateway computer, and / or an edge server. A network adapter or network interface in each computing / processing device receives the computer instructions from the network and forwards the computer instructions for storage in the computer-readable storage medium in each computing / processing device.
[0111] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present specification. In this regard, each block in the flowchart or block diagram may represent a module, a segment of a program, or a portion of computer instructions, which contains one or more executable computer instructions for implementing the specified logical function. In some alternative implementations, the functions noted in the blocks may occur in a different order than noted in the accompanying drawings. For example, two consecutive blocks may actually be executed substantially in parallel, or they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and combinations of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system that performs the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer instructions. As is well known to those skilled in the art, implementation by hardware, implementation by software, and implementation by a combination of software and hardware are equivalent.
[0112] The embodiments of the present specification have been described above. The above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope of the described embodiments. The choice of terms used herein is intended to best explain the principles of the embodiments, the practical application, or improvements to the technology in the market, or to enable other ordinary skilled persons in the art to understand the embodiments disclosed herein.
Claims
1. A blood pressure prediction method, characterized in that: include: Acquiring a photoplethysmography signal, an electrocardiogram signal, and a heart sound signal within a preset time length; Based on each cardiac cycle within the preset time length, determining the time at which a key feature point in the photoplethysmography signal is located, the time at which a key feature point in the electrocardiogram signal is located, and the time at which a key feature point in the heart sound signal is located; Determine the time difference between the key feature points in all pairwise combinations of signals in each cardiac cycle according to the time at which the key feature points in the photoplethysmography signal of each cardiac cycle, the time at which the key feature points in the electrocardiogram signal are located, and the time at which the key feature points in the heart sound signal are located; The current blood pressure value is predicted based on the time difference between the key characteristic points in all the pairwise combined signals in each cardiac cycle.
2. The method according to claim 1, characterized in that The key feature points in the photoplethysmography signal are the contraction peak-to-peak point and the valley point, the key feature points in the electrocardiogram signal are the wave peak point, and the key feature points in the heart sound signal are the first heart sound peak point and the second heart sound peak point.
3. The method according to claim 2, characterized in that Based on each cardiac cycle, the time difference between the key feature point in the photoplethysmography signal and the key feature point in the electrocardiogram signal is the time difference between the peak point and the peak-to-peak point of the contraction, and the time difference between the peak point and the valley point. The time difference between the key feature point in the photoplethysmography signal and the key feature point in the heart sound signal is the time difference between the first heart sound peak point and the systolic peak-peak point, the time difference between the first heart sound peak point and the valley point, the time difference between the second heart sound peak point and the systolic peak-peak point, and the time difference between the second heart sound peak point and the valley point. The time difference between the key feature point in the electrocardiogram signal and the key feature point in the heart sound signal is the time difference between the peak point and the first heart sound peak point, and the time difference between the peak point and the second heart sound peak point.
4. The method according to claim 1, characterized in that: The predicting of the current blood pressure value according to the time difference between the key characteristic points in all the signals in pairs in each cardiac cycle includes: Determine the statistical characteristic values of the time differences between the key characteristic points in all the signals in pairs in each cardiac cycle, wherein the statistical characteristic values are at least two of the mean, standard deviation, maximum value, minimum value, median value, kurtosis, skewness, and coefficient of variation; The current blood pressure value is predicted based on the statistical characteristic values of the time differences between the key characteristic points in all the pairwise combined signals and the corresponding regression model fitting coefficients.
5. The method according to claim 4, characterized in that The method further comprises: Obtaining statistical characteristic values of time differences between key characteristic points in all pairwise combinations of signals corresponding to historical cardiac cycles and actual blood pressure values corresponding to historical cardiac cycles; Based on the linear regression model, the fitting coefficients of each regression model are determined according to the statistical characteristic values of the time differences between key characteristic points in all pairwise combinations of signals corresponding to the historical cardiac cycles and the actual blood pressure values corresponding to the historical cardiac cycles.
6. The method according to claim 4, characterized in that The method further comprises: Determine the correlation coefficient between each statistical characteristic value of the time difference between key characteristic points in all pairwise combinations of signals and the blood pressure value; Based on all the signals in pairs, the statistical characteristic value corresponding to the correlation coefficient exceeding the preset threshold is obtained; wherein, The method of predicting the current blood pressure value based on the statistical characteristic values of the time differences between the key characteristic points in all the pairwise combined signals and the corresponding regression model fitting coefficients comprises: The current blood pressure value is predicted based on the statistical characteristic value corresponding to the correlation coefficient in all pairwise combinations of the signals exceeding the preset threshold and the corresponding regression model fitting coefficient.
7. The method according to any one of claims 1 to 6, characterized in that: Before determining the time at which the key feature points in the photoplethysmography signal, the time at which the key feature points in the electrocardiogram signal are located, and the time at which the key feature points in the heart sound signal are located based on each cardiac cycle within the preset time length, the method further includes: Normalization processing and filtering processing are respectively performed on the photoelectric volumetric pulse wave mapping method signal, electrocardiogram signal and heart sound signal within a preset time length to obtain processed photoelectric volumetric pulse wave mapping method signal, electrocardiogram signal and heart sound signal.
8. A blood pressure prediction device, characterized in that: include: An acquisition module, used to acquire a photoplethysmography signal, an electrocardiogram signal and a heart sound signal within a preset time length; A key feature point determination module, for determining the time at which the key feature point in the photoplethysmography signal, the time at which the key feature point in the electrocardiogram signal is located, and the time at which the key feature point in the heart sound signal is located based on each cardiac cycle within the preset time length; A time difference determination module, for determining the time difference between the key feature points in all pairwise combinations of signals in each cardiac cycle according to the time at which the key feature points in the photoplethysmography signal of each cardiac cycle, the time at which the key feature points in the electrocardiogram signal are located, and the time at which the key feature points in the heart sound signal are located; The blood pressure value prediction module predicts the current blood pressure value according to the time difference between the key characteristic points in all the pairwise combined signals in each cardiac cycle.
9. A blood pressure prediction device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and the computer program is used to control the processor to operate so as to execute the method according to any one of claims 1 to 7.
10. An electronic device, characterized in that: include: The blood pressure prediction device, the sensor for collecting photoplethysmography signals, the sensor for collecting electrocardiogram signals and the sensor for collecting heart sound signals as claimed in claim 8 or 9, wherein: The sensor for collecting photoplethysmography signals, the sensor for collecting electrocardiogram signals and the sensor for collecting heart sound signals are all connected to the blood pressure prediction device.