Blood pressure detection method and device based on photoelectric pulse wave signal, equipment and medium

Abstract

The application discloses a blood pressure detection method and device based on a photoelectric pulse wave signal, equipment and a medium, and relates to the technical field of wearable devices and health monitoring. The blood pressure detection method comprises the following steps: preprocessing the collected photoelectric pulse wave signal, extracting characteristic parameters, classifying the characteristic parameters into a basic class and a compensation class according to the correlation between each characteristic parameter and blood pressure data, fitting the characteristic parameters in the basic class and the characteristic parameters in the compensation class through a regression equation, respectively calculating the basic blood pressure and the compensation blood pressure according to the two types of fitted parameters, and correcting the basic blood pressure according to the deviation value of the two, so as to obtain the blood pressure detection value based on the photoelectric pulse wave signal. The blood pressure detection method realizes continuous blood pressure detection of a moving individual, the detection result is high in accuracy, and the safety of high-G load training of the individual is ensured.

Description

Technical Field

[0001] This invention relates to the field of wearable devices and health monitoring technology, and in particular to a method, apparatus, equipment and medium for blood pressure detection based on photoelectric pulse wave signals. Background Technology

[0002] Pilots may face serious physiological problems such as reduced blood flow to the head, decreased vision, and loss of consciousness during high G-load training. To ensure the health of pilots and the quality of training, it is especially important to accurately monitor vital signs such as arterial blood flow, pulse, and blood oxygen in the pilot's head. This can help trainees to identify problems in a timely manner and take countermeasures to avoid accidents.

[0003] Currently, both domestically and internationally, this is usually achieved by embedding intelligent, interconnected portable bio-information monitoring terminals into head-worn devices to monitor human vital signs. However, in actual use, existing products are not comfortable or natural to wear, and the monitoring accuracy is difficult to guarantee during exercise. Continuous blood pressure monitoring is particularly challenging.

[0004] Therefore, ensuring continuous and accurate monitoring of human blood pressure under high G-load exercise, in order to reduce the risks of high G-load training, is an urgent problem to be solved by current wearable monitoring devices. Summary of the Invention

[0005] This invention provides a method, device, equipment, and medium for blood pressure detection based on photoelectric pulse wave signals, in order to solve the problem of continuous and accurate monitoring of human blood pressure under high G-load training.

[0006] This invention is achieved through the following technical solution:

[0007] In a first aspect, the present invention provides a blood pressure detection method based on photoelectric pulse wave signals, comprising:

[0008] Acquire the photoelectric pulse wave signal of the user to be tested;

[0009] The photoelectric pulse wave signal is filtered and denoised to remove noise signals, power frequency interference signals and baseline drift, resulting in a clean photoelectric pulse wave signal.

[0010] Pulse wave feature points are extracted from the clean photoelectric pulse wave signal, and feature parameters are extracted from the pulse wave feature points;

[0011] The extracted feature parameters are categorized into basic and compensatory types based on their correlation with blood pressure.

[0012] The basic class feature parameters are fitted using a multiple regression equation to obtain the fitted parameters of the basic class feature parameters; the compensation class feature parameters are fitted using a univariate regression equation to obtain the fitted parameters of the compensation class feature parameters.

[0013] The baseline blood pressure is calculated based on the fitting parameters of the basic class feature parameters, and the compensated blood pressure is calculated based on the fitting parameters of the compensation class feature parameters.

[0014] The baseline blood pressure is corrected based on the deviation between the baseline blood pressure and the compensated blood pressure to obtain the measured blood pressure of the user to be tested.

[0015] This invention utilizes photoelectric pulse wave signal acquisition to detect blood pressure in individuals during exercise. The photoelectric pulse wave signal acquisition device can be integrated with a head-mounted wearable device, sensing the human body's pulse wave signal non-contactly. This avoids data discontinuity caused by movement between the body and the wearable device during exercise. Furthermore, a novel method for obtaining accurate voltage detection results based on photoelectric pulse wave signals is proposed. By analyzing the correlation between characteristic parameters and voltage data, the characteristic parameters are classified and fitted individually. The base voltage value is then corrected based on the predicted voltage value of the compensated category, ensuring the accuracy of the final detection result. This classification method can yield different results depending on the user, allowing for voltage detection based on individual differences, thus improving the accuracy of voltage detection values ​​and ensuring the safety of athletes, especially those undergoing high-G training.

[0016] In one embodiment, the step of filtering and denoising the photoelectric pulse wave signal to remove high-frequency noise, power frequency interference, and baseline drift to obtain a clean photoelectric pulse wave signal specifically includes:

[0017] The photoelectric pulse wave signal is digitally filtered to remove high-frequency noise signals and power frequency interference signals;

[0018] The photoelectric pulse wave signal after digital filtering is subjected to wavelet transform to filter out noise signals and baseline offset, resulting in a clean photoelectric pulse wave signal.

[0019] In one embodiment, the pulse wave feature points include: the maximum and minimum values ​​of the photoelectric pulse wave signal waveform, the first-order differential zero-crossing point of the photoelectric pulse wave signal waveform, the second-order differential zero-crossing point of the photoelectric pulse wave signal waveform, and the maximum and minimum values ​​of the photoelectric pulse wave signal waveform.

[0020] In one embodiment, the characteristic parameters include: total pulse wave area within one cycle, systolic cycle area, diastolic cycle area, main wave amplitude, dicrotic wave amplitude, descending isthmus amplitude, dicrotic wave amplitude, pulse wave area change, pulse wave cycle, pulse wave systolic cycle, pulse wave diastolic cycle, main wave slope, and cardiac output.

[0021] In one implementation, the extracted feature parameters are categorized into basic and compensatory classes based on their correlation with blood pressure, including:

[0022] Stable pulse wave data and blood pressure data are extracted from the MIMIC IV database, and feature parameters are extracted from the pulse wave data.

[0023] The Pearson correlation coefficient between each characteristic parameter of the pulse wave data and blood pressure is calculated. Characteristic parameters with a Pearson correlation coefficient greater than 0.7 are classified into the basic class, and characteristic parameters with a Pearson correlation coefficient in the range of 0.4 to 0.7 are classified into the compensation class.

[0024] In one implementation, the formula for calculating the baseline blood pressure is:

[0025]

[0026] Among them, BP m Indicates baseline blood pressure, M i Let β represent the i-th feature parameter in the base class, n represent the total number of feature parameters in the base class, and β represent the ith feature parameter in the base class. i Represents the characteristic parameter M i The fitting coefficients of the multiple regression equation, where a represents the fitting constant;

[0027] The formula for calculating the compensated blood pressure is:

[0028]

[0029] BP comp_j =α j C j +b j

[0030] Among them, BP comp Indicates compensation for blood pressure, ω j BP represents the weighting factor of the j-th feature parameter in the compensation class. comp_j C represents the compensated blood pressure corresponding to the j-th feature parameter in the compensation class. j Let α represent the j-th feature parameter in the compensation class. j b j These are the characteristic parameters C j The fitting coefficients and fitting constants of the univariate linear regression equation;

[0031] In one implementation, the step of correcting the baseline blood pressure based on the deviation between the baseline blood pressure and the compensated blood pressure to obtain the measured blood pressure of the user to be tested includes:

[0032] The deviation between baseline blood pressure and compensated blood pressure, based on experimental data and empirical values, is quantified using the following formula:

[0033]

[0034] Among them, BP gap This represents the deviation between the baseline blood pressure and the compensated blood pressure, expressed in BP. gap =BP m -BP comp ΔBP represents the quantized value of the deviation.

[0035] The baseline blood pressure is corrected based on the quantified value to obtain the measured blood pressure of the user to be tested, expressed by the following formula:

[0036] BP = BP m ±ΔBP

[0037] BP stands for blood pressure measurement. m The baseline blood pressure is represented by ΔBP, which is the quantified value of the deviation between the baseline blood pressure and the compensated blood pressure, ± based on BP. gap The sign of BP is determined if BP gap If BP is positive, then the addition is calculated. gap If the result is negative, then subtraction is performed.

[0038] A second aspect of the present invention provides a blood pressure detection method based on photoelectric pulse wave signals, comprising:

[0039] The acquisition module is used to acquire the photoelectric pulse wave signal of the user to be detected;

[0040] The preprocessing module is used to filter and reduce noise in the photoelectric pulse wave signal, removing noise signals, power frequency interference signals and baseline drift to obtain a clean photoelectric pulse wave signal.

[0041] The feature extraction module is used to extract pulse wave feature points from the clean photoelectric pulse wave signal and extract feature parameters through the pulse wave feature points.

[0042] The feature classification module is used to classify the extracted feature parameters into basic and compensatory categories based on the correlation between the feature parameters and blood pressure.

[0043] The feature fitting module is used to perform regression fitting on the basic class feature parameters using a multiple regression equation to obtain the fitting parameters of the basic class feature parameters; and to perform regression fitting on the compensation class feature parameters using univariate regression to obtain the fitting parameters of the compensation class feature parameters.

[0044] The blood pressure calculation module is used to calculate the baseline blood pressure based on the fitting parameters of the basic feature parameters and to calculate the compensated blood pressure based on the fitting parameters of the compensation feature parameters.

[0045] The blood pressure correction module is used to correct the baseline blood pressure based on the deviation between the baseline blood pressure and the compensated blood pressure, so as to obtain the test blood pressure of the user to be tested.

[0046] A third aspect of the present invention provides a head-mounted electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the blood pressure detection method based on photoelectric pulse wave signal as described in any of the first aspects of the present invention.

[0047] A third aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, characterized in that, when the computer program is executed by a processor, it implements the blood pressure detection method based on photoelectric pulse wave signal as described in any one of the first aspects of the present invention.

[0048] Compared with existing technologies, this invention has the following advantages and beneficial effects: The blood pressure detection method based on photoelectric pulse wave signals continuously monitors the blood pressure of individuals during exercise, without relying on the coupling degree between the wearable device and the individual. The characteristic parameters are classified and processed according to the correlation between pulse wave characteristic parameters and blood pressure signals. Baseline blood pressure is corrected by compensating for blood pressure, thereby enabling accurate blood pressure detection of individuals during exercise based on continuous photoelectric pulse wave signal acquisition results, ensuring the safety of personnel undergoing high-G load training. Attached Figure Description

[0049] To more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly described below. It should be understood that the following drawings only show some embodiments of the present invention and should not be considered as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort. In the drawings:

[0050] Figure 1 This is a flowchart of a blood pressure detection method based on photoelectric pulse wave signal according to an embodiment of the present invention;

[0051] Figure 2 This is a waveform diagram of an original photoelectric pulse wave signal according to an embodiment of the present invention;

[0052] Figure 3 This is a waveform diagram of a photoelectric pulse wave after wavelet denoising according to an embodiment of the present invention;

[0053] Figure 4A The coefficients of each layer of the photoelectric pulse wave signal after 9-layer wavelet decomposition are shown.

[0054] Figure 4B The results are shown after setting the wavelet decomposition detail coefficients to zero;

[0055] Figure 5 This is a schematic diagram of a pulse wave waveform and feature points according to an embodiment of the present invention;

[0056] Figure 6 This is a schematic diagram of some characteristic parameters of the pulse wave according to an embodiment of the present invention;

[0057] Figure 7 This is a model diagram of calculating blood pressure based on pulse wave characteristic parameters according to an embodiment of the present invention;

[0058] Figure 8 This is a comparison chart of systolic blood pressure (SBP) obtained by the method of this invention and Bland-Altman readings measured by a standard blood pressure device;

[0059] Figure 9 This is a comparison chart of diastolic blood pressure (DBP) obtained by the method of this invention and Bland-Altman readings measured by a standard blood pressure device; Detailed Implementation

[0060] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the embodiments and accompanying drawings. The illustrative embodiments and descriptions of the present invention are only used to explain the present invention and are not intended to limit the present invention.

[0061] It should be noted that the terms "comprising" and "having" and any variations thereof in the specification, claims, and accompanying drawings of this invention are intended to cover non-exclusive inclusion, for example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to other steps or units inherent in the device.

[0062] The terminology used in the various embodiments of the invention is for the purpose of describing particular embodiments only and is not intended to limit the various embodiments of the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which the various embodiments of the invention pertain. The terms (such as those defined in a generally used dictionary) are to be interpreted as having the same meaning as in the context of the relevant technical field and are not to be interpreted as having an idealized or overly formal meaning, unless clearly defined in the various embodiments of the invention.

[0063] To address the issues of existing wearable devices being uncomfortable to wear, difficult to continuously monitor blood pressure during exercise, and lacking in accuracy, this invention proposes a blood pressure detection method, device, equipment, and medium based on photoelectric pulse wave signals. The device continuously collects photoelectric pulse wave signals from the human body during movement using a head-mounted device, and obtains accurate blood pressure detection results through signal processing. This method is suitable for monitoring vital signs during exercise, and is particularly applicable to pilots during high-G training.

[0064] Figure 1 The diagram shows a flowchart of a blood pressure detection method based on photoelectric pulse wave signals according to an embodiment of the present invention. The method includes the following steps.

[0065] S1, acquire the photoelectric pulse wave signal of the user to be detected.

[0066] S2 performs filtering and noise reduction processing on the acquired photoelectric pulse wave signal to remove noise signals, power frequency interference signals and baseline drift, and obtain a clean photoelectric pulse wave signal.

[0067] S3 extracts pulse wave feature points from clean photoelectric pulse wave signals and extracts feature parameters through these pulse wave feature points.

[0068] S4 categorizes the extracted feature parameters into basic and compensatory categories based on their correlation with blood pressure.

[0069] S5. The basic class feature parameters are fitted by a multiple regression equation to obtain the fitted parameters of the basic class feature parameters; the compensation class feature parameters are fitted by a univariate regression equation to obtain the fitted parameters of the compensation class feature parameters.

[0070] S6 calculates the baseline blood pressure based on the basic class fitting parameters and the compensated blood pressure based on the compensated class fitting parameters.

[0071] S7 corrects the baseline blood pressure based on the deviation between the baseline blood pressure and the compensated blood pressure to obtain the test blood pressure of the user to be tested.

[0072] The voltage detection method of this invention embeds a portable photoelectric sensor bio-information monitoring terminal into a head-mounted wearable device to collect photoplethysmographic (PPG) signals from the user's forehead in a non-contact manner. Compared to contact sensors, even if the coupling position between the user's head and the head-mounted device changes during movement, it does not affect the continuous acquisition of signals, thereby avoiding the loss of monitoring data and ensuring the continuity of subsequent data processing.

[0073] Based on this, the present invention proposes an accurate blood pressure detection method for monitoring photoelectric pulse wave signals, so as to accurately predict blood pressure through continuous photoelectric pulse wave signals, thereby ensuring the continuity and accuracy of blood pressure monitoring.

[0074] Step S2 is the data preprocessing step. The signal acquired by the head-mounted device contains noise, which needs to be filtered to obtain a clean photoelectric pulse wave signal, thereby improving the accuracy of subsequent data processing. Preprocessing includes filtering high-frequency and low-frequency noise signals, eliminating power frequency interference signals, and baseline drift shifting.

[0075] In one implementation, the preprocessing step is as follows:

[0076] S2-1, First, the acquired photoelectric pulse wave signal is digitally filtered to remove high-frequency, low-frequency and power frequency interference signals;

[0077] S2-2, Wavelet transform is performed on the digitally filtered photoelectric pulse wave signal to further filter out noise signals and baseline offset, resulting in a clean photoelectric pulse wave signal.

[0078] Since the frequency of pulse wave signals is mainly concentrated in the range of 0.4Hz to 10Hz, high-frequency, low-frequency and power frequency interference can be filtered out through digital filtering. Then, wavelet transform is used to remove baseline drift and further filter noise signals, retaining the effective data in the photoelectric pulse wave signal.

[0079] like Figure 2 The image shows the waveform of the acquired raw photoelectric pulse wave signal. The above process performed the following steps on the raw signal: First, the signal was decomposed into 9 levels using the sym8 wavelet basis function, as shown below. Figure 3 The image shown is the waveform of the photoelectric pulse wave after wavelet denoising; then the coefficients of each layer were extracted, such as... Figure 4A As shown in the figure, the curves represent the coefficients of each level after the signal undergoes nine-level wavelet decomposition, used to analyze the signal's characteristics in different frequency bands; the detail coefficients are then set to zero to remove high-frequency noise and baseline drift, as shown below. Figure 4BAs shown; finally, the baseline signal and the signal after removing baseline drift are reconstructed, and the signal is filtered and denoised multiple times to obtain the final processed signal.

[0080] Step S3 is a feature extraction step for the clean photoelectric pulse wave signal. First, multiple pulse wave feature points are extracted based on the waveform of the photoelectric pulse wave signal, and then multiple pulse wave feature parameters are extracted based on the pulse wave feature points.

[0081] Among them, the pulse wave characteristic points include: the maximum and minimum values ​​of the photoelectric pulse wave signal waveform, the first-order differential zero-crossing point of the photoelectric pulse wave signal waveform, the second-order differential zero-crossing point of the photoelectric pulse wave signal waveform, and the maximum and minimum values ​​of the photoelectric pulse wave signal waveform.

[0082] Figure 5 The diagram shows a pulse wave waveform and its characteristic points. According to the waveform and mathematical meaning of the pulse wave, the peak and trough are points b and c, respectively. b, c, f, and g are extreme points, which are the 1st, 2nd, 3rd, and 4th zero-crossing points after the first-order difference of the waveform. d and e are the points where the concavity and convexity change, which are the zero-crossing points of the second-order difference between ce and e.

[0083] A single pulse wave cycle waveform can be divided by using adjacent pulse wave troughs (point b) or adjacent pulse wave peaks (point c).

[0084] Figure 6 The diagram shows a partial representation of the characteristic parameters of a pulse wave. The extracted characteristic parameters include: total pulse wave area S within one cycle, systolic cycle area S1, diastolic cycle area S2, main wave amplitude H1, dicrotic wave amplitude H2, descending isthmus amplitude H3, dicrotic wave amplitude H4, pulse wave area change K, pulse wave period T, pulse wave systolic period T1, pulse wave diastolic period T2, main wave slope V, and cardiac output Z = H(1 + T1 / T2).

[0085] Further, before extracting the characteristic parameters of the pulse wave, the waveform data is first normalized. The vertical distance H1 from the peak of the pulse wave time-domain signal to the signal baseline is denoted as 1. After normalization, the pulse wave amplitude parameters are: H2 / H1, H3 / H1, H4 / H1; the period normalization parameters are: T2 / T, T1 / T; and the area normalization parameters are S1 / S, S2 / S. Based on the change in pulse wave area, the waveform characteristic parameter K is derived, which represents the size of the pulsogram area of ​​the cardiac cycle.

[0086]

[0087] in P represents the average amplitude of the waveform over one period. s P represents the peak amplitude value. d The trough amplitude value is represented by the characteristic value K, which indicates the area of ​​the pulsogram during the cardiac cycle.

[0088] Before predicting blood pressure based on the extracted pulse wave feature parameters, the feature parameters are classified according to the correlation between each feature parameter and the blood pressure data.

[0089] In one implementation, the Pearson correlation coefficient between the characteristic parameters and blood pressure data is calculated to determine the strength of the correlation between each characteristic parameter and blood pressure. The calculation formula is as follows:

[0090]

[0091] Where, x i y i Let i represent the time series of feature parameters and blood pressure data, respectively, where i represents a time point. denoted as the mean values ​​of the feature parameters and blood pressure data, respectively, and r is the Pearson correlation coefficient between the feature parameters and blood pressure data.

[0092] The Pearson correlation coefficient ranges from -1 to 1, and its absolute value represents the magnitude of the correlation. The correlation coefficient is an important basis for feature classification. Specifically, step S4 is as follows:

[0093] S4-1, Obtain stable pulse wave data PPG and blood pressure data ABP from the historical monitoring data of the user to be tested, and extract feature parameters from the pulse wave data PPG according to the above embodiment;

[0094] S4-2, calculate the Pearson correlation coefficient between each feature parameter extracted in S4-1 and the blood pressure data. Feature parameters with a Pearson correlation coefficient greater than 0.7 are classified into the basic class M, and feature parameters with a Pearson correlation coefficient in the range of 0.4 to 0.7 are classified into the compensation class C.

[0095] First, PPG and APB signals were screened from the individual's historical monitoring data. The more stable signals were selected, and time-domain parameters and blood pressure values ​​were extracted. The correlation between pulse wave BP characteristic parameters and blood pressure was preliminarily analyzed using SPSS software. Table 1 shows the results of the correlation analysis between SBP / DBP and characteristic parameters.

[0096] Table 1. Results of Correlation Analysis between SBP / DBP and Characteristic Parameters

[0097]

[0098] Table 1 shows the correlation between systolic blood pressure (SBP), diastolic blood pressure (DBP), and various characteristic parameters of the pulse wave. Blood pressure changes are influenced by many factors, and further stepwise regression analysis is needed to determine the magnitude of the influence of these characteristic parameters on diastolic blood pressure (DBP) and systolic blood pressure (SBP).

[0099] In step S5, a multiple regression equation is used to perform regression fitting on the feature parameters in the basic class and the compensation class to obtain the fitting parameters of the basic class and the fitting parameters of the compensation class.

[0100] Multiple linear regression (MLR) is a common method in statistics used to analyze the linear relationship between two or more independent variables and a dependent variable. This method attempts to find a best-fit linear equation to describe the relationship between the independent and dependent variables, thus enabling the prediction or explanation of changes in the target variable. The basic form of an MLR model can be expressed as:

[0101] Y = a + β1M1 + β2M2 + ... + β n M n

[0102] Among them, M n This represents the parameters to be fitted, i.e., the independent variables. In this embodiment, it represents each feature parameter, β. n represents the fitting parameters for multiple linear regression, a is the fitting constant, and Y represents the fitting curve, i.e., the dependent variable.

[0103] Multiple linear regression (MLR) models can handle the relationship between multiple independent variables and a single dependent variable, making them applicable to various research fields and applications. The regression coefficients provided by the model can be interpreted as the average effect of a one-unit change in one independent variable on the dependent variable, while keeping other variables constant. This interpretability makes MLR highly valuable in practical applications, especially when understanding the relationships between variables is crucial. Compared to some complex nonlinear models or machine learning algorithms, MLR has relatively lower data requirements. It does not require large amounts of training data and is easier to process and analyze.

[0104] The univariate linear regression equation between compensation type C and blood pressure data is expressed as:

[0105]

[0106] Where, α 1～m and b 1～m Let C be the coefficient to be fitted. 1～m Here, m represents the feature parameters corresponding to the compensation type, and m is the total number of feature parameters in the compensation type.

[0107] Furthermore, in addition to fitting the parameters, the compensation calculation model also needs to calculate the weighting factors based on the correlation coefficient:

[0108]

[0109] Where, ω j r is the weighting factor for the j-th compensation type feature parameter.j Let be the correlation strength between the j-th compensation type feature parameter and the blood pressure data, and m be the total number of feature parameters in the compensation class.

[0110] Based on the correlation between different feature types and blood pressure, a weighted bias compensation amount is applied to achieve accurate detection. The compensation model is as follows: Figure 7 As shown, multiple linear regression is performed on the basic class to obtain the fitting parameters of the multiple regression equation, which are the fitting parameters of the basic class feature parameters. Weighted linear regression is performed on the compensation class to obtain the fitting parameters of the univariate linear regression equation, which are the fitting parameters of the compensation class feature parameters.

[0111] The baseline blood pressure is calculated based on the fitted parameters of the basic class feature parameters, using the following formula:

[0112]

[0113] Among them, BP m Indicates baseline blood pressure, M i Let β represent the i-th feature parameter in the base class, n represent the total number of feature parameters in the base class, and β represent the ith feature parameter in the base class. i Represents the characteristic parameter M i The fitting coefficients of the multiple regression equation, where a represents the fitting constant.

[0114] The compensating blood pressure is calculated based on the fitted parameters of the compensating feature parameters, using the following formula:

[0115]

[0116] BP comp_j =α j C j +b j

[0117] Among them, BP comp Indicates compensation for blood pressure, ω j BP represents the weighting factor of the j-th feature parameter in the compensation class. comp_j C represents the compensated blood pressure corresponding to the j-th feature parameter in the compensation class. j Let α represent the j-th feature parameter in the compensation class. j b j These are the characteristic parameters C j The fitting of the univariate linear regression equation and the fitting constant.

[0118] Further, in step S7, the baseline blood pressure is corrected based on the deviation between the baseline blood pressure and the compensated blood pressure to obtain the measured blood pressure of the user to be tested, specifically including:

[0119] S7-1 Quantifies the deviation between baseline blood pressure and compensated blood pressure based on empirical values ​​from experimental data. The formula is expressed as follows:

[0120]

[0121] Among them, BP gap This represents the deviation between the baseline blood pressure and the compensated blood pressure, expressed in BP. gap =BP m -BP comp ΔBP represents the quantized value of the deviation. 0, 2, 5, and 8 are empirical values ​​taken based on experimental data.

[0122] S7-2, the baseline blood pressure is corrected based on the quantified value of the deviation to obtain the measured blood pressure of the user to be tested. The formula is expressed as follows:

[0123] BP = BP m ±ΔBP

[0124] BP stands for blood pressure measurement. m BP represents the baseline blood pressure, and ΔBP represents the quantified deviation between the baseline blood pressure and the compensated blood pressure. ± is based on BP. gap The sign of BP is determined if BP gap If BP is positive, then the addition is calculated. gap If the result is negative, then subtraction is performed.

[0125] For different individuals, the influence of characteristic parameters on blood pressure varies. The absolute value of the correlation coefficient test, |r| = 0.40, is used as the threshold. The closer the correlation is to 1, the more significant it is and the better the fit.

[0126] Table 2 shows the correlation analysis results between systolic blood pressure (SBP) and characteristic parameters obtained from regression fitting. Table 3 shows the correlation analysis results between diastolic blood pressure (DBP) and characteristic parameters obtained from regression fitting. Table 4 shows the selection of characteristic parameters for different individuals (comp 1, comp 2, comp 3, comp 4, comp 5) of systolic blood pressure (SBP). Table 5 shows the selection of characteristic parameters for different individuals of diastolic blood pressure (DBP).

[0127] Table 2 Correlation Analysis of SBP and Characteristic Parameters

[0128]

[0129] Table 3 Correlation Analysis between DBP and Specific Parameters

[0130]

[0131]

[0132] Table 4. Selection of different individual characteristic parameters for SBP

[0133]

[0134] Table 5. Selection of different individual characteristic parameters for DBP

[0135]

[0136] Correlation analysis of Tables 2, 3, 4, and 5 shows that the influence of pulse wave characteristic parameters on blood pressure data varies among individuals. By using the characteristic parameter classification method of this invention to classify different characteristic parameters for each test subject, a separate blood pressure calculation model is obtained, improving the accuracy of individual blood pressure detection.

[0137] The blood pressure detection method based on photoplethysmography (PPG) wave signal proposed in this invention has the following beneficial effects: On the one hand, it can continuously monitor the user's blood pressure parameters without interfering with the user's normal activities and training, providing a reliable basis for the user's physical assessment; on the other hand, it also supports the monitoring of the user's real-time health status data, allowing the user to make better training arrangements based on their own blood pressure parameters, thus ensuring the user's physical health.

[0138] Using the Bland-Altman method, 70 groups of measurement data were selected for consistency analysis with actual blood pressure values, and compared with the blood pressure detection method of this invention. The comparison results are as follows: Figure 8 , Figure 9 As shown, the majority of the error scatter points are within ±5 mmHg. The blood pressure values ​​measured by the blood pressure detection method based on photoelectric pulse wave signals of this invention show good consistency with the results measured by electronic blood pressure monitors. The error ranges of SBP and DBP are both within the 96% confidence interval, indicating good consistency between the two methods and proving that the method of this invention has high accuracy.

[0139] A second aspect of the present invention provides a blood pressure detection device based on photoelectric pulse wave signals, comprising:

[0140] The acquisition module is used to acquire the photoelectric pulse wave signal of the user to be detected;

[0141] The preprocessing module is used to filter and reduce noise in the photoelectric pulse wave signal, removing noise signals, power frequency interference signals and baseline drift to obtain a clean photoelectric pulse wave signal.

[0142] The feature extraction module is used to extract pulse wave feature points from clean photoelectric pulse wave signals and extract feature parameters through the pulse wave feature points;

[0143] The feature classification module is used to classify the extracted feature parameters into basic and compensatory classes based on the correlation between the feature parameters and blood pressure.

[0144] The feature fitting module is used to perform regression fitting on the basic class feature parameters using a multiple regression equation to obtain the fitting parameters of the basic quantity class feature parameters; and to perform regression fitting on the compensation class feature parameters using univariate regression to obtain the fitting parameters of the compensation class feature parameters.

[0145] The blood pressure calculation module is used to calculate the baseline blood pressure based on the fitting parameters of the basic feature parameters and to calculate the compensated blood pressure based on the fitting parameters of the compensation feature parameters.

[0146] The blood pressure correction module is used to correct the baseline blood pressure based on the deviation between the baseline blood pressure and the compensated blood pressure, so as to obtain the measured blood pressure of the user to be tested.

[0147] Furthermore, the preprocessing module includes a digital filtering module and a wavelet transform module; the digital filtering module is used to perform digital filtering on the photoelectric pulse wave signal to filter out high-frequency, low-frequency and power frequency interference signals; the wavelet transform module is used to perform wavelet transform on the photoelectric pulse wave signal processed by the digital filtering module to filter out noise signals and baseline offset, and obtain a clean photoelectric pulse wave signal.

[0148] Furthermore, the pulse wave feature points extracted by the feature extraction module include: the maximum and minimum values ​​of the photoelectric pulse wave signal waveform, the first-order difference zero-crossing point of the photoelectric pulse wave signal waveform, the second-order difference zero-crossing point of the photoelectric pulse wave signal waveform, and the maximum and minimum values ​​of the photoelectric pulse wave signal waveform.

[0149] Furthermore, the feature extraction module extracts the following feature parameters within a cycle: total area of ​​pulse wave, area of ​​systolic cycle, area of ​​diastolic cycle, amplitude of main wave, amplitude of dichroic wave, amplitude of descending isthmus, amplitude of dichroic wave, change in pulse wave area, pulse wave cycle, pulse wave systolic cycle, pulse wave diastolic cycle, slope of main wave, and cardiac output.

[0150] Furthermore, the feature classification module calculates the Pearson correlation coefficient between each feature parameter of the pulse wave data and blood pressure, classifies feature parameters with a Pearson correlation coefficient greater than 0.7 into the basic class, and classifies feature parameters with a Pearson correlation coefficient in the range of 0.4 to 0.7 into the compensation class.

[0151] Furthermore, the blood pressure calculation module includes a baseline blood pressure calculation module and a compensatory blood pressure calculation module. The baseline blood pressure calculation module calculates the baseline blood pressure using the following formula:

[0152]

[0153] Among them, BP m Indicates baseline blood pressure, M i Let β represent the i-th feature parameter in the base class, n represent the total number of feature parameters in the base class, and β represent the ith feature parameter in the base class. i Represents the characteristic parameter Mi The fitting coefficients of the multiple regression equation, where a represents the fitting constant.

[0154] The compensating blood pressure calculation module calculates the compensating blood pressure using the following formula:

[0155]

[0156] BP comp_j =α j C j +b j

[0157] Among them, BP comp Indicates compensation for blood pressure, ω j BP represents the weighting factor of the j-th feature parameter in the compensation class. comp_j C represents the compensated blood pressure corresponding to the j-th feature parameter in the compensation class. j Let α represent the j-th feature parameter in the compensation class. j b j These are the characteristic parameters C j The fitting of the univariate linear regression equation and the fitting constant.

[0158] Furthermore, the blood pressure correction module includes a deviation value quantification module and a correction module. The deviation value quantification module calculates the deviation between the baseline blood pressure and the compensated blood pressure using the following formula:

[0159] BP gap =BP m -BP comp

[0160] The deviation value is then quantified using the following formula:

[0161]

[0162] Among them, BP gap This represents the deviation between the baseline blood pressure and the compensated blood pressure, expressed in BP. gap =BP m -BP comp ΔBP represents the quantized value of the deviation.

[0163] The correction module calculates blood pressure using the following correction formula:

[0164] BP = BP m ±ΔBP

[0165] BP stands for blood pressure measurement. m The baseline blood pressure is represented by ΔBP, which is the quantified value of the deviation between the baseline blood pressure and the compensated blood pressure, ± based on BP. gap The sign of BP is determined if BP gapIf BP is positive, then the addition is calculated. gap If the result is negative, then subtraction is performed.

[0166] A third aspect of the present invention provides a head-mounted electronic device, which includes a processor and a memory, wherein the number of processors may be one or more. The memory, as a computer-readable storage medium, can be used to store software programs, computer-executable programs, and modules. The processor executes various functional applications and data processing of the electronic device by running the software programs, instructions, and modules stored in the memory, thereby implementing the blood pressure detection method based on photoelectric pulse wave signals according to any of the above embodiments of the present invention.

[0167] The memory may primarily comprise a program storage area and a data storage area. The program storage area may store the operating system and at least one application program required for a given function; the data storage area may store data created based on terminal usage. Furthermore, the memory may include high-speed random access memory (RAM) and non-volatile memory, such as at least one disk storage device, flash memory, or other non-volatile solid-state storage device. In some instances, the memory may further include memory remotely located relative to the processor, which can be connected to the electronic device via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks (LANs), mobile communication networks, and combinations thereof.

[0168] Embodiments of the present invention also provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the blood pressure detection method based on photoelectric pulse wave signals according to any embodiment of the present invention.

[0169] The computer storage medium of this invention can be any combination of one or more computer-readable media. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. For example, a computer-readable storage medium can be, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this invention, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0170] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, capable of sending, propagating, or transmitting programs for use by or in connection with an instruction execution system, apparatus, or device.

[0171] Embodiments of the present invention also provide a computer program product that, when run on a computer, causes the computer to execute the blood pressure detection method based on photoelectric pulse wave signals according to any of the above embodiments of the present invention.

[0172] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A blood pressure detection method based on photoelectric pulse wave signals, characterized in that, include: Acquire the photoelectric pulse wave signal of the user to be tested; The photoelectric pulse wave signal is filtered and denoised to remove noise signals, power frequency interference signals and baseline drift, resulting in a clean photoelectric pulse wave signal. Pulse wave feature points are extracted from the clean photoelectric pulse wave signal, and feature parameters are extracted from the pulse wave feature points; The extracted feature parameters are categorized into basic and compensatory types based on their correlation with blood pressure. The basic class feature parameters are fitted using a multiple regression equation to obtain the fitted parameters of the basic class feature parameters; the compensation class feature parameters are fitted using a univariate regression equation to obtain the fitted parameters of the compensation class feature parameters. The baseline blood pressure is calculated based on the fitting parameters of the basic class feature parameters, and the compensated blood pressure is calculated based on the fitting parameters of the compensation class feature parameters. The baseline blood pressure is corrected based on the deviation between the baseline blood pressure and the compensated blood pressure to obtain the measured blood pressure of the user to be tested; The extracted feature parameters are categorized into basic and compensatory classes based on their correlation with blood pressure, including: Obtain stable pulse wave data and blood pressure data from the historical monitoring data of the user to be tested, and extract feature parameters from the pulse wave data; The Pearson correlation coefficient between each characteristic parameter of the pulse wave data and blood pressure is calculated. Characteristic parameters with a Pearson correlation coefficient greater than 0.7 are classified into the basic class, and characteristic parameters with a Pearson correlation coefficient in the range of 0.4 to 0.7 are classified into the compensation class.

2. The blood pressure detection method based on photoelectric pulse wave signal according to claim 1, characterized in that, The step of filtering and denoising the photoelectric pulse wave signal to remove high-frequency noise, power frequency interference, and baseline drift to obtain a clean photoelectric pulse wave signal specifically includes: The photoelectric pulse wave signal is digitally filtered to remove high-frequency, low-frequency, and power frequency interference signals; The photoelectric pulse wave signal after digital filtering is subjected to wavelet transform to filter out noise signals and baseline offset, resulting in a clean photoelectric pulse wave signal.

3. The blood pressure detection method based on photoelectric pulse wave signal according to claim 1, characterized in that, The pulse wave feature points include: the maximum and minimum values ​​of the photoelectric pulse wave signal waveform, the first-order differential zero-crossing point of the photoelectric pulse wave signal waveform, the second-order differential zero-crossing point of the photoelectric pulse wave signal waveform, and the maximum and minimum values ​​of the photoelectric pulse wave signal waveform.

4. The blood pressure detection method based on photoelectric pulse wave signal according to claim 3, characterized in that, The characteristic parameters include: total area of ​​pulse wave within one cycle, area of ​​systolic cycle, area of ​​diastolic cycle, amplitude of main wave, amplitude of descending isthmus, amplitude of dicrotic wave, change in pulse wave area, pulse wave cycle, pulse wave systolic cycle, pulse wave diastolic cycle, slope of main wave, and cardiac output.

5. The blood pressure detection method based on photoelectric pulse wave signal according to claim 1, characterized in that, The formula for calculating the baseline blood pressure is as follows: in, This indicates baseline blood pressure. Let represent the i-th feature parameter in the base class, and n represent the total number of feature parameters in the base class. Representing characteristic parameters The fit coefficients of the multiple regression equation. Represents the fitting constant; The formula for calculating the compensated blood pressure is: in, Indicates compensation for blood pressure. This represents the weighting factor of the j-th feature parameter in the compensation class. This represents the compensated blood pressure corresponding to the j-th feature parameter in the compensated class. This represents the j-th feature parameter in the compensation class. , These are the feature parameters. The fitting coefficients and fitting constants of the univariate linear regression equation.

6. The blood pressure detection method based on photoelectric pulse wave signal according to claim 5, characterized in that, The step of correcting the baseline blood pressure based on the deviation between the baseline blood pressure and the compensated blood pressure to obtain the measured blood pressure of the user to be tested includes: The deviation between the baseline blood pressure and the compensated blood pressure is quantified based on empirical values ​​from experimental data, expressed by the following formula: in, This represents the deviation between the baseline blood pressure and the compensated blood pressure. ; The quantized value representing the deviation value; The baseline blood pressure is corrected based on the quantified value to obtain the measured blood pressure of the user to be tested, expressed by the following formula: in, This indicates a blood pressure check. This indicates baseline blood pressure. This represents a quantified value indicating the deviation between the baseline blood pressure and the compensated blood pressure. according to The sign determines, if If it is positive, then calculate the addition; if If the result is negative, then subtraction is performed.

7. A blood pressure detection device based on photoelectric pulse wave signals, characterized in that, include: The acquisition module is used to acquire the photoelectric pulse wave signal of the user to be detected; The preprocessing module is used to filter and reduce noise in the photoelectric pulse wave signal, removing noise signals, power frequency interference signals and baseline drift to obtain a clean photoelectric pulse wave signal. The feature extraction module is used to extract pulse wave feature points from the clean photoelectric pulse wave signal and extract feature parameters through the pulse wave feature points. The feature classification module is used to classify the extracted feature parameters into basic and compensatory categories based on the correlation between the feature parameters and blood pressure. The feature fitting module is used to perform regression fitting on the basic class feature parameters using a multiple regression equation to obtain the fitting parameters of the basic class feature parameters; and to perform regression fitting on the compensation class feature parameters using univariate regression to obtain the fitting parameters of the compensation class feature parameters. The blood pressure calculation module is used to calculate the baseline blood pressure based on the fitting parameters of the basic feature parameters and to calculate the compensated blood pressure based on the fitting parameters of the compensation feature parameters. A blood pressure correction module is used to correct the baseline blood pressure based on the deviation between the baseline blood pressure and the compensated blood pressure, so as to obtain the test blood pressure of the user to be tested. The extracted feature parameters are categorized into basic and compensatory classes based on their correlation with blood pressure, including: Obtain stable pulse wave data and blood pressure data from the historical monitoring data of the user to be tested, and extract feature parameters from the pulse wave data; The Pearson correlation coefficient between each characteristic parameter of the pulse wave data and blood pressure is calculated. Characteristic parameters with a Pearson correlation coefficient greater than 0.7 are classified into the basic class, and characteristic parameters with a Pearson correlation coefficient in the range of 0.4 to 0.7 are classified into the compensation class.

8. A head-mounted electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the blood pressure detection method based on photoelectric pulse wave signal as described in any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the blood pressure detection method based on photoelectric pulse wave signal as described in any one of claims 1-6.

Images