Blood pressure compensation measurement methods, devices, and smartwatches for sports scenarios

Abstract

This invention relates to a blood pressure compensation measurement method, device, and smartwatch for sports scenarios. The method generates a lactate accumulation marker by fusing motion acceleration and heart rate signals. Based on this marker, specific frequency components of the pressure oscillation wave signal are extracted and a time-domain envelope is generated. The variation in the width of the envelope's main peak and the abnormal frequency-domain energy distribution are quantified to generate a tension anomaly, which is then mapped to a dynamic vascular volume change curve. A biomechanical compensation curve is constructed using a static vascular volume benchmark, and the envelope, free from vascular tension interference, is reconstructed through frequency-domain convolution. Finally, blood pressure feature points are located based on the differential features of the reconstructed envelope, and a compensated blood pressure value is output. This method establishes, for the first time, a compensation model for lactate metabolism and vascular biomechanical characteristics, solving the problem of systematic shift of feature points caused by abnormal vascular tension after exercise, and significantly reducing blood pressure measurement errors.

Description

Technical Field

[0001] This invention relates to the field of blood pressure compensation measurement technology, specifically to a blood pressure compensation measurement method, device, and smartwatch for sports scenarios. Background Technology

[0002] In the field of health monitoring, blood pressure measurement during exercise is crucial for assessing cardiovascular function and exercise safety. Traditional oscillometric blood pressure measurement relies on identifying characteristic points of the cuff pressure oscillation wave, determining systolic and diastolic blood pressure by analyzing the points of maximum slope and curvature change of the envelope. This method has good reliability at rest and has been widely used in home blood pressure monitors and medical devices. With the development of wearable devices, the demand for real-time blood pressure monitoring during exercise is increasing, especially in scenarios such as fitness training and athlete physiological monitoring. Accurately obtaining post-exercise blood pressure data is vital for preventing cardiovascular events.

[0003] However, existing blood pressure measurement technologies face significant limitations in post-exercise scenarios. During anaerobic exercise, lactic acid buildup in muscle tissue triggers abnormal changes in vascular tension, leading to distortions in the envelope morphology of the cuff pressure oscillation wave. This physiological interference manifests as feature point shift, envelope broadening, and abnormal spectral energy distribution, making it impossible for traditional algorithms to accurately locate blood pressure feature points. While existing solutions attempt to suppress limb movement interference through motion artifact filtering or acceleration compensation, they fail to address the core issue of altered vascular biomechanical properties caused by lactic acid metabolism, resulting in systematic biases in measurement results. More critically, existing technologies lack the ability to quantitatively model dynamic changes in vascular tension, failing to establish a compensation mechanism between lactic acid buildup and blood pressure measurement distortion, posing a risk of misleading blood pressure data obtained by users after exercise. Summary of the Invention

[0004] Based on this, the purpose of the present invention is to provide a blood pressure compensation measurement method, device, and smartwatch that can accurately eliminate the interference of lactic acid accumulation in sports scenarios.

[0005] The objective of this invention is achieved through the following solution:

[0006] In a first aspect, the present invention provides a blood pressure compensation measurement method for sports scenarios, comprising the following steps:

[0007] S1: Process the motion acceleration signal collected by the triaxial accelerometer and the heart rate signal collected by the heart rate sensor, calculate the motion intensity index based on acceleration and the heart rate recovery slope based on the heart rate change rate, and generate a lactate accumulation indicator based on the preset intensity threshold and the preset recovery threshold.

[0008] S2: Process the cuff pressure oscillation wave signal collected by the pressure sensor, extract the intrinsic mode functions in the 0.5-5Hz frequency band and perform Hilbert transform to generate the time domain envelope;

[0009] S3: Process the time-domain envelope, calculate the change in the width of the main peak of the envelope relative to the static reference and the energy ratio of the 8-12Hz band to the 1-4Hz band, and generate the tension abnormality degree characterizing vascular tension abnormality.

[0010] S4: Process the tension abnormality, calculate the vasodilation amount obtained through the nonlinear mapping function, and construct the vascular volume compensation curve by combining it with the exponential decay time function.

[0011] S5: Process the time-domain envelope and vascular volume compensation curve, reconstruct the compensated envelope through frequency-domain convolution operation, locate the maximum slope point and the second-order positive change point on the reconstructed envelope, and output the compensated blood pressure value to eliminate motion interference.

[0012] In one embodiment, S1 of the blood pressure compensation measurement method for a sports scenario provided by the present invention specifically includes the following steps:

[0013] S11: Perform cube root integration on the motion acceleration signal acquired by the triaxial accelerometer, calculate the integral value of the cube root of the triaxial composite acceleration within the time window, and generate the motion intensity index.

[0014] S12: Perform linear regression processing on the heart rate signal collected by the heart rate sensor, calculate the slope value of the heart rate decrease rate after exercise stops, and generate the heart rate recovery slope.

[0015] S13: Perform dual threshold logic judgment on exercise intensity index and heart rate recovery slope. When the exercise intensity exceeds the preset intensity threshold and the heart rate recovery rate is lower than the preset recovery threshold, generate an active lactic acid accumulation flag.

[0016] In one embodiment, step S2 of the blood pressure compensation measurement method for a sports scenario provided by the present invention specifically includes the following steps:

[0017] S21: Perform empirical mode decomposition on the cuff pressure oscillation wave signal acquired by the pressure sensor, decompose the signal into multiple intrinsic mode function components, and extract the intrinsic mode functions in the 0.5-5Hz frequency band.

[0018] S22: Perform Hilbert transform on the intrinsic mode functions, calculate the instantaneous amplitude of the analytic signal, and generate the initial envelope;

[0019] S23: The initial envelope is smoothed by filtering, and high-frequency fluctuation interference is eliminated by calculating the mean within the time window to generate an optimized time-domain envelope.

[0020] In one embodiment, S3 of the blood pressure compensation measurement method for sports scenarios provided by the present invention specifically includes the following steps:

[0021] S31: Perform main peak detection processing on the time domain envelope, locate the global maximum point of the envelope and measure its half-height width, and generate the main peak width variation;

[0022] S32: Perform frequency domain energy analysis on the time domain envelope, calculate the energy distribution ratio of the 8-12Hz band to the 1-4Hz band, and generate the frequency domain characteristic ratio.

[0023] S33: The ratio of the main peak width change to the frequency domain feature is weighted and fused, and the comprehensive offset is calculated by combining the static benchmark value to generate the tension anomaly degree. The tension anomaly degree is used to indicate the degree of vascular tension abnormality caused by lactic acid accumulation.

[0024] In one embodiment, the formula for calculating the degree of tension abnormality in a blood pressure compensation measurement method for motion scenarios provided by the present invention is as follows:

[0025] ;

[0026] in, The degree of vascular tension abnormality is characterized by the extent of abnormal vascular tension. This refers to the change in the width of the main peak. The static reference half-height width, The frequency domain characteristic ratio, For time-domain feature weighting factors, This is the frequency domain feature weighting factor.

[0027] In one embodiment, S4 of the blood pressure compensation measurement method for sports scenarios provided by the present invention specifically includes the following steps:

[0028] S41: The tension abnormality is processed by a nonlinear function mapping. The tension abnormality is converted into the proportion of vascular volume change through hyperbolic tangent transformation to generate vascular diastole.

[0029] S42: The vasodilation volume is modeled by exponential decay, and the time decay function is used to simulate the dynamic recovery process of vascular volume during lactate metabolism, and a volume change curve that decreases over time is constructed.

[0030] S43: Perform baseline overlay processing on the volume change curve, add the volume change curve to the preset static vascular volume reference value to generate a vascular volume compensation curve, which is used for subsequent frequency domain filtering to eliminate abnormal vascular tension interference.

[0031] In one embodiment, S5 of the blood pressure compensation measurement method for sports scenarios provided by the present invention specifically includes the following steps:

[0032] S51: Perform frequency domain convolution processing on the time domain envelope and the blood vessel volume compensation curve. By multiplying the envelope spectrum with the transfer function of the compensation filter in the frequency domain and performing an inverse Fourier transform to convert it to the time domain, a reconstructed compensation envelope is generated.

[0033] S52: Perform differential feature extraction on the reconstructed compensation envelope, find the maximum point by calculating the first derivative and detect the position point where the second derivative first turns from negative to positive, and generate characteristic moments of systolic blood pressure and diastolic blood pressure.

[0034] S53: Perform pressure value mapping processing on the characteristic time of systolic blood pressure and characteristic time of diastolic blood pressure, query the cuff pressure value at the corresponding time according to the characteristic time, and generate a compensated blood pressure value that includes compensated systolic blood pressure value and compensated diastolic blood pressure value.

[0035] In one embodiment, S52 of the blood pressure compensation measurement method for a sports scenario provided by the present invention specifically includes the following steps:

[0036] S521: Perform first-order derivative calculation on the reconstructed compensation envelope, calculate the instantaneous slope value of each sampling point on the envelope using the central difference method, and generate a first-order derivative sequence;

[0037] S522: Perform extreme point detection processing on the first derivative sequence, scan the entire sequence to find the maximum positive value point and record its corresponding time, and generate the characteristic time of systolic blood pressure.

[0038] S523: Perform second derivative calculation on the reconstructed compensation envelope, calculate the curvature change value of each sampling point on the envelope using the second-order central difference method, and search for the first point where the curvature changes from negative to positive after the systolic pressure characteristic moment to generate the diastolic pressure characteristic moment.

[0039] Secondly, the present invention provides a blood pressure compensation measurement device for sports scenarios, the device being configured with the following modules:

[0040] The lactate accumulation marker generation module is used to process the motion acceleration signal collected by the triaxial accelerometer and the heart rate signal collected by the heart rate sensor, calculate the exercise intensity index based on acceleration and the heart rate recovery slope based on the heart rate change rate, and generate a lactate accumulation marker based on a preset intensity threshold and a preset recovery threshold.

[0041] The time-domain envelope generation module is used to process the cuff pressure oscillation wave signal collected by the pressure sensor, extract the intrinsic mode functions in the 0.5-5Hz frequency band and perform Hilbert transform to generate the time-domain envelope.

[0042] The tension anomaly calculation module is used to process the time-domain envelope, calculate the change in the width of the main peak of the envelope relative to the static reference and the energy ratio of the 8-12Hz frequency band to the 1-4Hz frequency band, and generate the tension anomaly degree characterizing vascular tension abnormality.

[0043] The vascular volume compensation curve construction module is used to process tension abnormality, calculate the vascular dilation amount obtained through a nonlinear mapping function, and construct the vascular volume compensation curve by combining it with an exponential decay time function.

[0044] The compensated blood pressure output module is used to process the time-domain envelope and the vascular volume compensation curve. It reconstructs the compensated envelope through frequency-domain convolution operations and locates the maximum slope point and the second-order positive change point on the reconstructed envelope, and outputs the compensated blood pressure value to eliminate motion interference.

[0045] Thirdly, this application provides a smartwatch, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the blood pressure compensation measurement method for any of the above-mentioned sports scenarios.

[0046] In summary, the blood pressure compensation measurement method for motion scenarios provided in this application is based on a dual-threshold judgment mechanism of the cube root integral of acceleration signal and heart rate recovery slope. This mechanism can accurately identify physiological state changes caused by lactic acid accumulation, overcoming the misjudgment defects caused by traditional methods relying solely on a single motion sensor. Secondly, by extracting specific frequency band components of pressure oscillation waves and combining them with Hilbert envelope extraction technology, signal distortion caused by abnormal vascular tension can be effectively separated, solving the problem of feature waveform submersion under motion interference. Furthermore, the innovative fusion of temporal envelope peak broadening features and frequency domain energy distribution anomaly features to construct tension anomaly degree enables quantitative characterization of changes in vascular biomechanical properties. Most importantly, the vascular volume dynamic response model established through nonlinear mapping and exponential decay function can accurately simulate the recovery law of vascular tension during lactic acid metabolism. Combined with frequency domain convolution compensation technology to reconstruct the envelope, it can completely eliminate the systematic offset of blood pressure feature points caused by changes in vascular volume. Finally, the use of a differential feature joint detection mechanism in the positioning stage ensures stable capture of real blood pressure feature points under complex interference environments. This method establishes for the first time a complete technology chain from lactate metabolism kinetics to blood pressure measurement compensation, which can maintain millimeter-mercury level measurement accuracy in scenarios with abnormal vascular tension after exercise, and solves the core pain point of inaccurate blood pressure monitoring by wearable devices in scenarios such as fitness rehabilitation and sports training.

[0047] To better understand and implement this invention, the following detailed description is provided in conjunction with the accompanying drawings. Attached Figure Description

[0048] Figure 1A flowchart illustrating a blood pressure compensation measurement method in a sports scenario provided in an embodiment of this application;

[0049] Figure 2 A schematic diagram of the process for generating tension anomalies provided in an embodiment of this application;

[0050] Figure 3 A schematic diagram of the process for generating a compensated blood pressure value that includes a compensated systolic blood pressure value and a compensated diastolic blood pressure value, provided for an embodiment of this application;

[0051] Figure 4 This is a schematic diagram of the structure of a blood pressure compensation measurement device for a sports scenario, provided as another embodiment of this application. Detailed Implementation

[0052] To facilitate understanding of the present invention, a more complete description will be given below with reference to the accompanying drawings. Preferred embodiments of the invention are shown in the drawings. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to provide a thorough and complete understanding of the disclosure of the invention.

[0053] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0054] In one embodiment, such as Figure 1 As shown, a blood pressure compensation measurement method for sports scenarios is provided. This embodiment illustrates the application of this method to a terminal. It is understood that this method can also be applied to a server, or to a device including both a terminal and a server, and implemented through interaction between the terminal and the server. This embodiment uses a smartwatch with an integrated blood pressure measurement module as a carrier. The hardware includes a three-axis accelerometer, an optical heart rate sensor, a piezoresistive pressure sensor, and an embedded processor. The software is ported to the embedded system after algorithm verification. In this embodiment, the method includes the following steps:

[0055] S1: Process the motion acceleration signal collected by the triaxial accelerometer and the heart rate signal collected by the heart rate sensor, calculate the motion intensity index based on acceleration and the heart rate recovery slope based on the heart rate change rate, and generate a lactic acid accumulation indicator based on the preset intensity threshold and the preset recovery threshold.

[0056] Specifically, the system integrates a three-axis accelerometer into the smartwatch to collect motion acceleration signals, and simultaneously collects heart rate signals through an optical heart rate sensor integrated into the smartwatch. The motion acceleration signals collected by the smartwatch are filtered to remove limb tremor noise; the heart rate signals collected by the smartwatch are adaptively filtered to eliminate motion artifacts, wherein the reference signal for adaptive filtering is taken from the output of the three-axis accelerometer integrated into the smartwatch.

[0057] Preferably, the system performs vector synthesis on the preprocessed triaxial acceleration signal, and then uses a sliding window to calculate the average value of the synthesized signal to obtain an acceleration-based motion intensity index. The system determines a preset intensity threshold through clinical trials. During the clinical trials, the system simultaneously acquires the subject's blood lactate concentration data through a monitoring unit connected to a smartwatch, and determines the preset intensity threshold based on the correlation between blood lactate concentration and the motion intensity index.

[0058] The system identifies the moment of exercise cessation by detecting changes in acceleration signals. When the exercise intensity index drops from above a certain level to below a certain level, the system determines that exercise has stopped. The system extracts heart rate sequences collected by a smartwatch over a period of time after the moment of exercise cessation, performs linear fitting on the heart rate sequences, and obtains the heart rate recovery slope. The system determines a preset recovery threshold based on clinical data and the correlation between the heart rate recovery slope and lactate metabolism status. When both the exercise intensity index and the heart rate recovery slope reach the preset recovery threshold, the system sets the lactate accumulation flag to a specific state; otherwise, the system sets the lactate accumulation flag to another state.

[0059] S2: Process the cuff pressure oscillation wave signal acquired by the pressure sensor, extract the intrinsic mode functions in the 0.5-5Hz frequency band and perform Hilbert transform to generate the time domain envelope.

[0060] Specifically, the system controls the piezoresistive pressure sensor of the smartwatch to collect pressure oscillation wave signals during the inflation and deflation of the cuff. The sensor's measurement range covers the entire range of cuff pressure changes. The system performs Butterworth high-pass filtering on the raw pressure oscillation wave signal to remove baseline drift caused by slow cuff deflation, and then performs low-pass filtering to eliminate high-frequency electromagnetic interference, resulting in a preprocessed signal. The system performs empirical mode decomposition on the preprocessed signal, repeatedly calculating the upper and lower envelopes of the signal and taking the average as the trend term. The trend term is then removed from the raw signal until the remaining signal meets the intrinsic mode function (EMF) determination criteria, thus separating out multiple EMFs.

[0061] For example, the system filters out the frequency bands corresponding to the intrinsic components of the blood pressure oscillation wave, excluding respiratory interference and muscle tremor interference. Respiratory interference is defined as a frequency band lower than the intrinsic component frequency band, while muscle tremor interference is defined as a frequency band higher than the intrinsic component frequency band. Further, the system performs Hilbert transform on the filtered intrinsic mode functions to construct analytic signals with the real part being the original intrinsic mode functions and the imaginary part being the result of the Hilbert transform. The system calculates the magnitude of each analytic signal and takes the arithmetic mean of all instantaneous amplitudes to obtain the time-domain envelope characterizing the amplitude variation of the cuff pressure oscillation wave.

[0062] S3: Process the time-domain envelope, calculate the change in the width of the main peak of the envelope relative to the static reference and the energy ratio of the 8-12Hz band to the 1-4Hz band, and generate the tension abnormality degree characterizing vascular tension abnormality.

[0063] Specifically, the system repeatedly performs blood pressure measurements on the user at rest. After each measurement, it extracts the time-domain envelope, calculates the half-width at half-maximum (WHM) of the envelope's main peak, and takes the average of the multiple WHM measurements as the static baseline for the width of the main peak. In motion scenarios, the system calculates the first derivative of the real-time time-domain envelope, identifies the extreme point where the first derivative changes from positive to negative to pinpoint the location of the main peak, calculates the real-time WHM of the main peak, and obtains the change relative to the static baseline through difference and ratio calculations. The system performs a Fast Fourier Transform on the preprocessed pressure oscillation wave signal to obtain the frequency domain distribution, and then divides it into two characteristic frequency bands. The first frequency band corresponds to the high-frequency vibration of vascular smooth muscle, i.e., the energy change when tension is abnormal, and the second frequency band corresponds to the low-frequency vibration of basic blood flow. The system performs integral calculations on the two frequency bands to obtain energy values, obtains energy ratios through ratio calculations, and constructs a weighted summation calculation model based on the change and the energy ratio. The weight coefficients are calibrated by multiple linear regression of multiple sets of blood pressure measurement samples after exercise. Substituting the change and the energy ratio into the model, the tension abnormality degree, which characterizes the degree of vascular tension abnormality, is obtained. The range of tension abnormality degree values ​​corresponds to different abnormality levels, and the higher the level, the stronger the compensation processing is required.

[0064] S4: Process the tension abnormality, calculate the vasodilation amount obtained through the nonlinear mapping function, and construct the vascular volume compensation curve by combining it with the exponential decay time function.

[0065] Specifically, this embodiment establishes the correlation between abnormal vascular tension and vascular diastole through experiments. Ultrasonic Doppler equipment is used to measure changes in the arterial diameter of subjects after exercise, and the abnormal vascular tension is recorded simultaneously to form an experimental database. A fitting operation is performed on the database data to obtain a nonlinear mapping function. This function includes parameters corresponding to the maximum vascular diastole of the radial artery at the wrist of a healthy adult, and parameters ensuring that the diastole approaches the maximum diastole when the abnormal tension reaches its upper limit. Based on the above processing, the system substitutes the real-time abnormal tension into the function to obtain the additional vascular diastole caused by lactic acid buildup.

[0066] Specifically, the system defines an exponential decay time function based on the time dependence of lactate metabolism. The function includes the initial decay time constant of the lactate metabolism rate at the moment of exercise cessation and the decay coefficient determined based on the experimental measurement results of lactate half-life. The input is the time after exercise cessation, and the output is the decay coefficient reflecting the recovery trend of vasodilation. The system also introduces a frequency parameter corresponding to the periodic characteristics of the cuff deflation rate. Through multiplication, the vasodilation, the output value of the exponential decay time function, and the value of the frequency-related periodic function are combined to construct a vascular volume compensation curve that dynamically adjusts with time, which is used to offset the influence of abnormal vascular volume on the time-domain envelope.

[0067] S5: Process the time-domain envelope and vascular volume compensation curve, reconstruct the compensated envelope through frequency-domain convolution operation, locate the maximum slope point and the second-order positive change point on the reconstructed envelope, and output the compensated blood pressure value to eliminate motion interference.

[0068] Specifically, the system performs Fast Fourier Transform (FFT) on the time-domain envelope and the vascular volume compensation curve, converting the time-domain signals into frequency-domain signals, resulting in the frequency-domain signals corresponding to the time-domain envelope and the vascular volume compensation curve. The system then performs a product operation on the two frequency-domain signals in the frequency domain, which is equivalent to a convolution operation in the time domain, avoiding edge effects that occur during time-domain convolution. Finally, the system performs an Inverse Fast Fourier Transform (IFFT) on the frequency-domain signals after the product operation, converting them back to the time domain to obtain the reconstructed compensation envelope.

[0069] Furthermore, the system performs a first-order derivative operation on the compensation envelope to obtain its first-order derivative signal. Then, it locates the maximum value point within this signal, which is the systolic blood pressure characteristic point, corresponding to the peak value of the rate of change of the oscillating wave amplitude when the cuff pressure and arterial systolic blood pressure are balanced. The system records the static cuff pressure at this point as the systolic blood pressure. The system then performs a second-order derivative operation on the compensation envelope to obtain its second-order derivative signal. Next, it locates the zero-crossing point in this signal, where the signal changes from negative to positive. This zero-crossing point is the diastolic blood pressure characteristic point, corresponding to the inflection point of the rate of change of the oscillating wave amplitude when the cuff pressure is lower than the arterial diastolic blood pressure. The system records the static cuff pressure at this point as the diastolic blood pressure. Finally, the systolic and diastolic blood pressures are integrated to generate a compensated blood pressure value. This compensated blood pressure value effectively eliminates motion interference, providing users with more accurate and reliable blood pressure measurement results, thus achieving the purpose of compensated blood pressure measurement in motion scenarios.

[0070] In summary, the blood pressure compensation measurement method for motion scenarios provided in this application is based on a dual-threshold judgment mechanism of the cube root integral of acceleration signal and heart rate recovery slope. This mechanism can accurately identify physiological state changes caused by lactic acid accumulation, overcoming the misjudgment defects caused by traditional methods relying solely on a single motion sensor. Secondly, by extracting specific frequency band components of pressure oscillation waves and combining them with Hilbert envelope extraction technology, signal distortion caused by abnormal vascular tension can be effectively separated, solving the problem of feature waveform submersion under motion interference. Furthermore, the innovative fusion of temporal envelope peak broadening features and frequency domain energy distribution anomaly features to construct tension anomaly degree enables quantitative characterization of changes in vascular biomechanical properties. Most importantly, the vascular volume dynamic response model established through nonlinear mapping and exponential decay function can accurately simulate the recovery law of vascular tension during lactic acid metabolism. Combined with frequency domain convolution compensation technology to reconstruct the envelope, it can completely eliminate the systematic offset of blood pressure feature points caused by changes in vascular volume. Finally, the use of a differential feature joint detection mechanism in the positioning stage ensures stable capture of real blood pressure feature points under complex interference environments. This method establishes for the first time a complete technology chain from lactate metabolism kinetics to blood pressure measurement compensation, which can maintain millimeter-mercury level measurement accuracy in scenarios with abnormal vascular tension after exercise, and solves the core pain point of inaccurate blood pressure monitoring by wearable devices in scenarios such as fitness rehabilitation and sports training.

[0071] In one embodiment, S1 of the blood pressure compensation measurement method for a sports scenario provided by the present invention specifically includes the following steps:

[0072] S11: Perform cube root integration on the motion acceleration signal acquired by the triaxial accelerometer, calculate the integral value of the cube root of the triaxial composite acceleration within the time window, and generate the motion intensity index.

[0073] Specifically, the system acquires motion acceleration signals via a smartwatch. The smartwatch's integrated three-axis accelerometer establishes a real-time data link with the embedded processor. The sensor acquires motion acceleration signals along the X, Y, and Z axes corresponding to the wearing position and transmits them directly to the processor. The system preprocesses the three-axis acceleration signals acquired by the smartwatch, using a filtering algorithm to remove limb tremor noise. This noise, generated by unintentional wrist shaking, interferes with the validity of the acceleration signal. After filtering, only signal components related to motion intensity are retained. After preprocessing, the system performs vector synthesis on the three-axis acceleration signals. Vector synthesis integrates the signals from the three axes into a single composite acceleration signal, avoiding misjudgments of intensity caused by changes in the direction of motion in a single axis signal, and ensuring that the composite signal reflects the overall motion amplitude.

[0074] The system performs a cube root operation on the synthesized acceleration signal. This operation compresses the numerical range of the synthesized acceleration signal while preserving the signal differences under different motion intensities. This prevents excessive signal amplification at high acceleration values ​​and signal masking at low acceleration values, providing a stable signal basis for subsequent integration calculations. The system sets a fixed time window, the length of which is determined based on the signal variation period under the motion state, ensuring that sufficient samples can be collected within the window to reflect the current motion intensity. Within the set time window, the system performs an integration operation on the synthesized acceleration signal after the cube root operation. The integration result is the motion intensity index, which is positively correlated with the motion intensity.

[0075] S12: Perform linear regression processing on the heart rate signal collected by the heart rate sensor, calculate the slope value of the heart rate decrease rate after exercise stops, and generate the heart rate recovery slope.

[0076] Specifically, the optical heart rate sensor integrated into the smartwatch emits a specific wavelength light signal. After the light signal penetrates the skin of the wrist, the sensor receives the light reflection signal generated by blood flow and converts it into a heart rate electrical signal. This electrical signal is transmitted to the smartwatch's embedded processor in real time. Preferably, the system preprocesses the heart rate electrical signal collected by the smartwatch. An adaptive filtering algorithm can be used to eliminate motion artifacts. Motion artifacts are caused by light reflection interference due to the relative displacement between the wrist skin and the sensor, and muscle contraction, which can cause false fluctuations in the heart rate signal. During filtering, the motion signal collected by the smartwatch's three-axis accelerometer is used as a reference, and the filtering parameters are adjusted in real time to remove artifacts and obtain a signal that reflects the true heart rate.

[0077] For example, the system identifies the moment of exercise cessation based on the exercise intensity index. When the exercise intensity index drops from above a certain dynamic threshold to below another dynamic threshold, and this state is maintained for a certain period of time, the system determines that exercise has stopped, records and marks the time node of the cessation. After exercise stops, the system extracts the heart rate signal within a specific time period after the cessation moment from the smartwatch's local storage unit, forming a continuous heart rate sequence. This time period needs to cover the initial stage of the heart rate decline from the post-exercise peak to ensure that the sequence can reflect the initial trend of heart rate recovery. The system performs linear regression processing on the heart rate sequence with time as the independent variable and heart rate as the dependent variable, establishes a linear regression model, calculates the slope value of the model to obtain the heart rate recovery slope, and a negative slope value indicates a decrease in heart rate, while the absolute value reflects the recovery rate.

[0078] S13: Perform dual threshold logic judgment on exercise intensity index and heart rate recovery slope. When the exercise intensity exceeds the preset intensity threshold and the heart rate recovery rate is lower than the preset recovery threshold, generate an active lactic acid accumulation flag.

[0079] Specifically, on the one hand, the system collects exercise intensity indices at different exercise intensities through a smartwatch, and simultaneously monitors changes in the subject's blood lactate concentration, recording the blood lactate concentration value corresponding to each exercise intensity index. When the exercise intensity index reaches a certain value, the blood lactate concentration exceeds the baseline concentration at rest and continues to rise; this value is determined as the preset intensity threshold. On the other hand, the system collects the heart rate recovery slope of different subjects after exercise stops through a smartwatch, and simultaneously monitors the corresponding blood lactate metabolism status. When the heart rate recovery slope is lower than a certain value, the blood lactate metabolism rate is significantly reduced, and the accumulation state is difficult to alleviate quickly; this value is determined as the preset recovery threshold.

[0080] For example, the system acquires the generated real-time exercise intensity index and compares it with a preset intensity threshold to determine whether it exceeds the threshold; simultaneously, it acquires the generated heart rate recovery slope and compares it with a preset recovery threshold to determine whether it is below the threshold. When the exercise intensity index exceeds the preset intensity threshold and the heart rate recovery slope is below the preset recovery threshold, the system determines that lactic acid accumulation exists and generates an activated lactic acid accumulation flag; if neither condition is met, the system determines that there is no significant lactic acid accumulation and generates an inactive lactic acid accumulation flag.

[0081] In one embodiment, step S2 of the blood pressure compensation measurement method for a sports scenario provided by the present invention specifically includes the following steps:

[0082] S21: Perform empirical mode decomposition on the cuff pressure oscillation wave signal acquired by the pressure sensor, decompose the signal into multiple intrinsic mode function components, and extract the intrinsic mode functions in the 0.5-5Hz frequency band.

[0083] Specifically, the system controls a piezoresistive pressure sensor inside the cuff to collect pressure oscillation wave signals during the cuff inflation and deflation process. The sensor's measurement range covers the entire range of cuff pressure changes, and the signal is transmitted to the system's main control unit after being output by the sensor. Further, the system preprocesses the original pressure oscillation wave signal, eliminating baseline drift and high-frequency electromagnetic interference caused by slow cuff deflation through filtering, resulting in a preprocessed signal. Simultaneously, the system performs empirical mode decomposition on the preprocessed signal. First, it calculates the upper and lower envelopes of the signal, then uses an interpolation algorithm to fit the extreme points to generate the upper and lower envelopes. Next, it takes the mean of the upper and lower envelopes as a trend term and removes this trend term from the original signal. The system repeats this filtering process until the remaining signal meets the intrinsic mode function (IMF) criteria, i.e., the number of extreme points is equal to or differs by one from the number of zero-crossing points. Finally, the preprocessed signal is decomposed into multiple IMF components. Based on the frequency band characteristics of the intrinsic components of the blood pressure oscillation wave, the system selects the components corresponding to the 0.5-5Hz frequency band from the multiple intrinsic mode function components obtained by decomposition. This frequency band is the main energy distribution range of the blood pressure oscillation wave, which can eliminate respiratory interference and muscle tremor interference, and uses it as the target signal for subsequent processing.

[0084] S22: Perform Hilbert transform on the intrinsic mode functions, calculate the instantaneous amplitude of the analytic signal, and generate the initial envelope.

[0085] Specifically, the system retrieves the filtered intrinsic mode function (IMF) data for the 0.5-5Hz frequency band, performs a Hilbert transform on the IMF component, and then performs a Fast Fourier Transform (FFT) on the time-series data of the IMF component to convert the time-domain signal into a frequency-domain signal. During the conversion, the system pads zero values ​​according to the data length of the IMF component to ensure that the frequency domain resolution can cover all frequency points within the 0.5-5Hz band and avoid frequency information loss. Next, the converted frequency-domain signal is multiplied by a sign function, the value of which is determined by the sign of the frequency value—1 when the frequency value is greater than zero and -1 when the frequency value is less than zero. This operation adjusts the phase characteristics of the frequency-domain signal so that the subsequent inverse transform can obtain a result that conforms to the definition of the Hilbert transform. Finally, an inverse Fast Fourier Transform (IFFT) is performed on the multiplied frequency-domain signal to convert the frequency-domain signal back into a time-domain signal, which is the Hilbert transform result of the IMF component.

[0086] Furthermore, the system uses the time-series data of the original IMF components as the real part and the corresponding Hilbert transform result as the imaginary part, combining them to form an analytic signal—each time point of the analytic signal contains both real and imaginary values. The system performs instantaneous amplitude calculation on the analytic signal: first, it calculates the square of the real part and the square of the imaginary part of the analytic signal at each time point, then adds the two squares to obtain a sum of squares; finally, it performs a square root operation on the sum of squares, and the result is the instantaneous amplitude at that time point. The instantaneous amplitude reflects the magnitude of the cuff pressure oscillation wave at the corresponding time point and is directly related to the intensity of the arterial pulsation. The system arranges the instantaneous amplitudes of each time point sequentially according to the time sequence of the signals collected by the smartwatch, forming continuous time-series data, which is the initial envelope.

[0087] S23: The initial envelope is smoothed by filtering, and high-frequency fluctuation interference is eliminated by calculating the mean within the time window to generate an optimized time-domain envelope.

[0088] Specifically, the system acquires the initial envelope time-series data. Through analysis of the time-series changes in the data, it is found that the initial envelope contains high-frequency fluctuation interference. The interference sources include two aspects: first, during the cuff deflation process, a slight control error in the deflation valve causes a slight fluctuation in cuff pressure, which is transmitted to the pressure sensor, forming pressure fluctuation interference; second, the electronic noise of the pressure sensor itself is generated by the current change in the internal circuit of the sensor, manifesting as high-frequency random signal fluctuations. These interferences cause irregular jitter in the shape of the initial envelope. If it is directly used for subsequent feature calculations, it will lead to deviations in the calculation results. Therefore, the system performs smoothing filtering on the initial envelope, using the mean value within the time window for filtering.

[0089] For example, the system analyzes the temporal variation period of the initial envelope, determining the variation period of the envelope by statistically analyzing the duration of the main peak of the initial envelope. A fixed time window duration is set based on the variation period, which must meet two conditions: first, it must be less than half the period of the main peak of the initial envelope to avoid excessive smoothing of the main peak shape and loss of key features due to an excessively long window; second, it must be greater than the period of high-frequency fluctuation interference to ensure that each window contains multiple high-frequency fluctuation periods, allowing fluctuation interference to be canceled out through averaging. The system divides the temporal data of the initial envelope into multiple consecutive windows according to the chronological order, with a fixed step size sliding time window. Each window contains a fixed number of temporal data points. For windows at the edges, if the number of data points is insufficient for the set window size, the system uses repeated filling of edge data points to ensure that the number of data points in the edge windows reaches the set value.

[0090] The system calculates the mean of all data points within each time window, sums the values ​​of all data points within the window, divides the sum by the number of data points in the window to obtain the mean of the window, and replaces the values ​​of all original data points within the window with this mean, completing the smoothing process for one window. The system repeats the above sliding window, mean calculation, and data replacement operations until all time-series data points of the initial envelope have been smoothed, resulting in an optimized time-domain envelope. The optimized time-domain envelope retains the main morphological features related to arterial pulsation, such as the primary and secondary peaks, in the initial envelope, while eliminating high-frequency fluctuation interference, making the changes in time-series data more closely match the true amplitude changes of arterial pulsation.

[0091] In one embodiment, such as Figure 2 As shown, step S3 of the blood pressure compensation measurement method for sports scenarios provided by the present invention specifically includes the following steps:

[0092] S31: Perform main peak detection processing on the time domain envelope, locate the global maximum point of the envelope and measure its half-height width, and generate the main peak width variation.

[0093] Specifically, the system first performs baseline drift correction on the time-domain envelope, and then constructs a baseline model using a linear fitting algorithm, as shown in the formula:

[0094] ;

[0095] Where B(t) is the baseline value at time t, k is the baseline slope (determined by the linear decrease in cuff deflation pressure), b is the initial baseline value at t=0, and t is the time variable. The system calculates the baseline values ​​at each time point using this formula, subtracts the corresponding baseline value from the original envelope data, and performs drift correction to ensure that the envelope only reflects arterial pulsation characteristics. After correction, the system performs main peak detection: traversing the time-series data points of the envelope, it locates the global maximum value point with the largest amplitude. , The time coordinate of the maximum value point. For the corresponding amplitude, calculate the half-height value. ,Towards Traverse forward and backward, locate the amplitude equal to Given the left endpoint t1 and the right endpoint t2, calculate the half-height width:

[0096] ;

[0097] Where W is the current half-height width of the main peak. The time at the left end of half the height, The time is the right endpoint of the half-height. The system calls the previously stored static baseline half-height width. When the lactic acid accumulation indicator is not activated, the smartwatch collects resting signals and obtains them through the same process, calculating the change in the width of the main peak:

[0098] ;

[0099] where ΔW is the change in the width of the main peak, W is the current full width at half maximum, is the static reference full width at half maximum. This change is correlated with the change in vascular tone, and the data are all from the collection and calculation of the smartwatch.

[0100] S32: Perform frequency-domain energy analysis on the time-domain envelope, calculate the ratio of the energy distribution between the 8 - 12 Hz frequency band and the 1 - 4 Hz frequency band, and generate a frequency-domain characteristic ratio.

[0101] Specifically, the system performs a discrete Fourier transform on the time-domain envelope, converting it into a frequency-domain signal. The formula is:

[0102] ;

[0103] where X(k) is the frequency-domain amplitude at the k-th frequency point, is the n-th time-domain envelope data point, N is the total number of data points, n is the time-domain index (0 ≤ n < N), k is the frequency-domain index (0 ≤ k < N), and j is the imaginary unit. The system completes the time-domain to frequency-domain conversion through this formula to obtain the amplitudes at each frequency point. Subsequently, the system calculates the power spectral density:

[0104] ;

[0105] where, is the power spectral density at the k-th frequency point ∣X(k)∣ is the modulus of X(k), N is the total number of data points, is the frequency value at the k-th frequency point. The system performs energy integration on the 8 - 12 Hz and 1 - 4 Hz frequency bands respectively according to the preset frequency bands:

[0106] ;

[0107] where E is the total energy of the target frequency band, is the starting frequency of the frequency band, is the ending frequency of the frequency band, is the power spectral density function. Calculate the energy of the 8 - 12 Hz frequency band and the energy

[0108] ;

[0109] where R is the frequency-domain characteristic ratio, The total energy in the 8-12Hz frequency band, This represents the total energy in the 1-4Hz frequency band, and changes in the ratio reflect the energy shift caused by abnormal vascular tension.

[0110] S33: The ratio of the main peak width change to the frequency domain feature is weighted and fused, and the comprehensive offset is calculated by combining the static benchmark value to generate the tension anomaly degree. The tension anomaly degree is used to indicate the degree of vascular tension abnormality caused by lactic acid accumulation.

[0111] Specifically, the system calls the previously stored static baseline values, including the static baseline half-height and width. Ratio to resting frequency domain characteristics When the lactate accumulation marker is inactive, the smartwatch collects the resting signal and obtains it through the same process, serving as a benchmark. The system performs weighted fusion on ΔW and R, first determining the time-domain feature weighting factor. Frequency domain feature weighting factor Both were obtained through fitting clinical data, and the system was substituted into the optimization formula to calculate the degree of tension abnormality:

[0112] ;

[0113] in, The degree of vascular tension abnormality is characterized by the extent of abnormal vascular tension. This refers to the change in the width of the main peak. The static reference half-height width, The frequency domain characteristic ratio, For time-domain feature weighting factors, This is the frequency domain feature weighting factor.

[0114] In one embodiment, S4 of the blood pressure compensation measurement method for sports scenarios provided by the present invention specifically includes the following steps:

[0115] S41: The tension abnormality is processed by a nonlinear function mapping. The tension abnormality is converted into the proportion of vascular volume change through hyperbolic tangent transformation, and the vascular diastolic volume is generated.

[0116] Specifically, the system uses nonlinear mapping to... To convert the change in vascular volume into a quantifiable proportion, the hyperbolic tangent transform is chosen as the mapping function—this function can... The range of any numerical value is compressed to the [0,1] interval to avoid extreme values. The value causes a proportional overflow of volume change, while retaining It exhibits a positive correlation with volume change. The mapping process can be achieved using the following formula:

[0117] ;

[0118] in, It is the vasodilation rate (i.e., the proportion of change in vascular volume, which is dimensionless). The maximum vascular volume change ratio is denoted as Σ(k) (based on measured vascular volume at rest, reflecting the upper limit of vascular dilation; data sourced from vascular-related signals collected by a smartwatch at rest), and k is the mapping adjustment coefficient (obtained through fitting clinical data to ensure...). When taking the median value It can linearly reflect changes in vascular tension. The tension anomaly degree generated in step 1.

[0119] The system accesses the data stored on the smartwatch. Substituting the parameter k into the equation yields... ΔV is calculated using the above formula. When When it increases, Approaching 1, Approaching When Dt decreases, Approaching 0, Approaching 0, it achieves a precise mapping from abnormal tension to vasodilation. All computational data comes from the results of prior collection and processing by the smartwatch.

[0120] S42: The vasodilation volume is modeled using exponential decay, and the time decay function is applied to simulate the dynamic recovery process of vascular volume during lactate metabolism, constructing a volume change curve that decreases over time.

[0121] Specifically, the system uses the generated vasodilation volume as a basis to simulate the dynamic recovery of vascular volume during lactic acid metabolism. Lactic acid metabolism gradually weakens over time, vascular tension gradually returns to normal, and the corresponding percentage change in vascular volume must decrease over time. Preferably, the system can use an exponential decay function for modeling, and the modeling process uses the following formula:

[0122] ;

[0123] in, Let λ be the dynamic volume change at time t, and λ be the decay coefficient, determined based on the lactate metabolism rate. This coefficient is obtained by fitting the relationship between lactate concentration and vascular volume changes at different time points. The data is also linked to the time of movement cessation recorded by a smartwatch, where t represents the time variable after movement stops. The system obtains the t value in real time through the smartwatch and substitutes it into the above formula to calculate the corresponding values ​​at different times t. The volume change curve is obtained by arranging the curves according to the time series. This curve exhibits an exponential decreasing trend as t increases: in the initial stage of motion cessation... Smaller Approaching 1, near As t increases, Increase Approaching 0, Vdecay(t) approaches 0, accurately simulating the dynamic recovery process of vascular volume caused by lactic acid metabolism.

[0124] S43: Perform baseline overlay processing on the volume change curve, add the volume change curve to the preset static vascular volume reference value to generate a vascular volume compensation curve, which is used for subsequent frequency domain filtering to eliminate abnormal vascular tension interference.

[0125] Specifically, the system needs to introduce a static vascular volume baseline value to ensure that the compensation curve can reflect the absolute change in vascular volume. This baseline value... The reference value for vascular volume at rest is obtained by the smartwatch when the lactate buildup indicator is inactive, by collecting the resting cuff pressure oscillation wave signal, combining it with a vascular volume estimation algorithm, and storing it in the smartwatch's local storage unit. Preferably, the compensation curve can be constructed using the following formula:

[0126] in, The curve for vascular volume compensation at time t is dimensionless. This is the baseline value for static vascular volume. This represents the generated dynamic volume change. The formula combines the dynamic volume change with a static benchmark through baseline overlay, enabling... It can reflect both the baseline vascular volume at rest and the dynamic changes in lactic acid metabolism after exercise.

[0127] The system accesses the data stored on the smartwatch. Combined with real-time computing Substituting into the above formula, we get The trend of the generated vascular volume compensation curve over time is similar to... Consistent, initial moment As t increases, it gradually approaches the value. This curve is used for subsequent frequency domain filtering to provide a quantitative basis for eliminating the interference of abnormal vascular tension on blood pressure measurement.

[0128] In one embodiment, such as Figure 3 As shown, step S5 of the blood pressure compensation measurement method for sports scenarios provided by the present invention specifically includes the following steps:

[0129] S51: Perform frequency domain convolution processing on the time domain envelope and the blood vessel volume compensation curve. By multiplying the envelope spectrum with the transfer function of the compensation filter in the frequency domain and performing an inverse Fourier transform to convert it to the time domain, a reconstructed compensation envelope is generated.

[0130] Specifically, the system calls the optimized time-domain envelope and the vascular volume compensation curve to ensure that the time axes of the two curves are synchronized. The system performs a discrete Fourier transform on the optimized time-domain envelope, converting the time-domain signal into a frequency-domain spectrum to obtain the frequency component distribution of the envelope; at the same time, it performs frequency-domain analysis on the vascular volume compensation curve to extract its corresponding compensation filter transfer function. This function is determined by the correlation between vascular volume changes and the frequency-domain response of blood pressure oscillation waves, and is used to adjust the envelope spectrum in a targeted manner.

[0131] The system performs element-wise multiplication of the envelope spectrum with the transfer function of the compensation filter in the frequency domain. This operation adjusts the amplitude of each frequency component of the envelope to counteract frequency domain distortion caused by abnormal vascular tension. The system then performs an inverse discrete Fourier transform on the resulting frequency domain signal, converting it back to the time domain. This yields a continuous curve that eliminates lactic acid buildup interference and closely approximates the resting state. This curve is defined as the reconstructed compensation envelope.

[0132] S52: Perform differential feature extraction on the reconstructed compensation envelope, find the maximum point by calculating the first derivative and detect the position point where the second derivative first turns from negative to positive, and generate the characteristic time of systolic blood pressure and diastolic blood pressure.

[0133] Specifically, the system performs differential feature extraction processing on the reconstructed compensation envelope and uses a sliding window algorithm to calculate the first derivative of the envelope. The system iterates through all data points of the first derivative, locates the data point with the largest amplitude, and defines this point as the moment when the amplitude of the reconstructed compensation envelope changes the fastest. This moment is defined as the systolic blood pressure characteristic moment. The system continues to calculate the second derivative of the first derivative, again using a sliding window algorithm to ensure calculation accuracy; the system iterates through the data points of the second derivative, detects the first position point where the value changes from negative to positive, and defines the moment corresponding to this position point as the diastolic blood pressure characteristic moment. Preferably, the systolic blood pressure characteristic moment and the diastolic blood pressure characteristic moment are obtained through the following steps:

[0134] S521: Perform first-order derivative calculation on the reconstructed compensation envelope, calculate the instantaneous slope value of each sampling point on the envelope using the central difference method, and generate a first-order derivative sequence.

[0135] Specifically, the system first calls the reconstructed compensation envelope data in the cache unit to determine the sampling interval of the data. This interval is consistent with the sampling frequency of the smartwatch's pressure sensor, ensuring the accuracy of the time dimension in subsequent differential calculations. Preferably, the system can use the central difference method to perform first-order derivative calculations on the reconstructed compensation envelope. For each sampling point on the envelope, two adjacent sampling points are selected. The instantaneous slope value of that point is obtained by dividing the difference in envelope amplitude between the two sampling points by twice the sampling interval. During the calculation process, for the sampling points at both ends of the envelope, the system uses the one-sided difference method to supplement the calculation. A sampling point following the first point and a sampling point preceding the last point are selected, and the amplitude difference is calculated with the first and last points respectively. This difference is then divided by the sampling interval to ensure that each sampling point in the entire envelope time sequence corresponds to an instantaneous slope value. The instantaneous slope values ​​of all sampling points are arranged in chronological order to form a first-order derivative sequence.

[0136] S522: Perform extreme point detection processing on the first derivative sequence, scan the entire sequence to find the maximum positive value point and record its corresponding time, and generate the characteristic time of systolic blood pressure.

[0137] Specifically, the system traverses each sampling point in the first derivative sequence in chronological order. Starting from the initial sampling point of the sequence, it sequentially extracts the instantaneous slope value of each sampling point, compares it with the instantaneous slope values ​​of adjacent sampling points, and records the maximum value and corresponding timestamp during the current traversal. The traversal process continues until the last sampling point of the first derivative sequence. The system finally determines the maximum positive value point in the entire sequence. The instantaneous slope value at this point is the maximum value of all sampling points in the sequence, and the value is positive, reflecting that the envelope has the fastest rate of ascent at that moment. The system extracts the timestamp corresponding to this maximum positive value point, which is the characteristic moment of systolic blood pressure, representing the time node when cuff pressure and arterial systolic blood pressure are in equilibrium.

[0138] S523: Perform second derivative calculation on the reconstructed compensation envelope, calculate the curvature change value of each sampling point on the envelope using the second-order central difference method, and search for the first point where the curvature changes from negative to positive after the systolic pressure characteristic moment to generate the diastolic pressure characteristic moment.

[0139] Specifically, the system again calls the generated reconstructed compensation envelope and uses the second-order central difference method to calculate the curvature change value of each sampling point. For the nth sampling point on the envelope, its curvature change value is obtained by dividing the difference between the first derivative of the (n+1)th sampling point and the first derivative of the (n-1)th sampling point by twice the sampling interval. The system determines the search range to be all sampling points after the systolic blood pressure characteristic moment, excluding the envelope rising phase before the systolic blood pressure characteristic moment, focusing on the critical change in the amplitude decreasing phase, and traversing the curvature change values ​​within this range. The system detects the first sampling point whose value changes from negative to positive, extracts the timestamp corresponding to the sampling point, and defines the timestamp as the diastolic blood pressure characteristic moment. Here, negative values ​​correspond to an accelerated decrease in the envelope amplitude, and positive values ​​correspond to a decelerated decrease in the amplitude. The transition point is the critical moment when the rate of decrease slows down, which matches the physiological process of the cuff pressure being lower than the diastolic blood pressure.

[0140] S53: Perform pressure value mapping processing on the characteristic time of systolic blood pressure and characteristic time of diastolic blood pressure, query the cuff pressure value at the corresponding time according to the characteristic time, and generate a compensated blood pressure value that includes compensated systolic blood pressure value and compensated diastolic blood pressure value.

[0141] Specifically, the system synchronously acquires the static pressure signal of the cuff collected by a piezoresistive pressure sensor. This signal records the static pressure value at different moments during the cuff's inflation and deflation. Based on the characteristic moment of systolic blood pressure, the system queries the cuff static pressure signal for the corresponding static pressure value at that moment; this pressure value is the compensated systolic blood pressure value after eliminating motion interference. Similarly, based on the characteristic moment of diastolic blood pressure, the system queries the cuff static pressure signal for the corresponding static pressure value at that moment; this pressure value is the compensated diastolic blood pressure value after eliminating motion interference. The system integrates the compensated systolic and compensated diastolic blood pressure values ​​into a data set, forming a compensated blood pressure value containing two parameters. This blood pressure value can be directly used for output display, providing accurate data for blood pressure monitoring in exercise scenarios.

[0142] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0143] Based on the same inventive concept, this application also provides a blood pressure compensation measurement device for implementing the blood pressure compensation measurement method for the above-mentioned exercise scenario. The solution provided by this device is similar to the implementation described in the above method. Therefore, the specific limitations of one or more embodiments of the blood pressure compensation measurement device for exercise scenarios provided below can be found in the limitations of the blood pressure compensation measurement method for exercise scenarios described above, and will not be repeated here.

[0144] Preferably, such as Figure 4 As shown, the present invention provides a blood pressure compensation measurement device 600 for sports scenarios, which is configured with the following modules:

[0145] The lactate accumulation marker generation module 610 is used to process the motion acceleration signal collected by the triaxial accelerometer and the heart rate signal collected by the heart rate sensor, calculate the exercise intensity index based on acceleration and the heart rate recovery slope based on the heart rate change rate, and generate a lactate accumulation marker based on a preset intensity threshold and a preset recovery threshold.

[0146] The time-domain envelope generation module 620 is used to process the cuff pressure oscillation wave signal collected by the pressure sensor, extract the intrinsic mode functions in the 0.5-5Hz frequency band and perform Hilbert transform to generate the time-domain envelope.

[0147] The tension anomaly calculation module 630 is used to process the time domain envelope, calculate the change in the width of the main peak of the envelope relative to the static reference and the energy ratio of the 8-12Hz frequency band to the 1-4Hz frequency band, and generate the tension anomaly degree characterizing vascular tension abnormality.

[0148] The vascular volume compensation curve construction module 640 is used to process the tension abnormality, calculate the vascular dilation amount obtained through the nonlinear mapping function, and construct the vascular volume compensation curve by combining it with the exponential decay time function.

[0149] The compensated blood pressure output module 650 is used to process the time-domain envelope and the vascular volume compensation curve. It reconstructs the compensated envelope through frequency-domain convolution operation and locates the maximum slope point and the second-order positive change point on the reconstructed envelope, and outputs the compensated blood pressure value to eliminate motion interference.

[0150] Preferably, the lactic acid accumulation marker generation module 610 provided in this application is configured with the following units:

[0151] The motion intensity index calculation unit is used to perform cube root integration on the motion acceleration signal collected by the triaxial accelerometer, calculate the integral value of the cube root of the triaxial composite acceleration within the time window, and generate the motion intensity index.

[0152] The heart rate recovery slope calculation unit is used to perform linear regression processing on the heart rate signal collected by the heart rate sensor, calculate the slope value of the rate of heart rate decline after exercise cessation, and generate the heart rate recovery slope.

[0153] The lactate accumulation flag determination unit is used to perform dual threshold logic judgment on exercise intensity index and heart rate recovery slope. When the exercise intensity exceeds the preset intensity threshold and the heart rate recovery rate is lower than the preset recovery threshold, an active lactate accumulation flag is generated.

[0154] Preferably, the time-domain envelope generation module 620 provided in this application is configured with the following units:

[0155] The intrinsic mode function extraction unit is used to perform empirical mode decomposition processing on the cuff pressure oscillation wave signal collected by the pressure sensor, decompose the signal into multiple intrinsic mode function components, and extract the intrinsic mode functions in the 0.5-5Hz frequency band.

[0156] The initial envelope generation unit is used to perform Hilbert transform processing on the intrinsic mode functions of the 0.5-5Hz frequency band, calculate the instantaneous amplitude of the analytic signal, and generate the initial envelope.

[0157] The envelope smoothing optimization unit is used to perform smoothing filtering on the initial envelope, eliminate high-frequency fluctuation interference by calculating the mean within the time window, and generate an optimized time-domain envelope.

[0158] Preferably, the tension anomaly calculation module 630 provided in this application is configured with the following units:

[0159] The main peak width detection unit is used to perform main peak detection processing on the time domain envelope, locate the global maximum point of the envelope and measure its half-height width, and generate the main peak width change.

[0160] The frequency domain energy ratio calculation unit is used to perform frequency domain energy analysis on the time domain envelope, calculate the energy distribution ratio of the 8-12Hz band and the 1-4Hz band, and generate frequency domain characteristic ratios.

[0161] The tension anomaly fusion calculation unit is used to perform weighted fusion processing on the change in the main peak width and the frequency domain feature ratio, and calculate the comprehensive offset by combining the static reference value to generate the tension anomaly. The tension anomaly is used to indicate the degree of vascular tension abnormality caused by lactic acid accumulation.

[0162] Preferably, the vascular volume compensation curve construction module 640 provided in this application is configured with the following units:

[0163] The vasodilation mapping unit is used to perform nonlinear function mapping on the tension abnormality. It converts the tension abnormality into the proportion of vascular volume change through hyperbolic tangent transformation to generate vasodilation.

[0164] The volume change curve modeling unit is used to perform exponential decay modeling of vasodilation. It applies a time decay function to simulate the dynamic recovery process of vascular volume during lactate metabolism and constructs a volume change curve that decreases over time.

[0165] The vascular volume compensation curve generation unit is used to perform baseline superposition processing on the volume change curve, adding the volume change curve to a preset static vascular volume reference value to generate a vascular volume compensation curve. The vascular volume compensation curve is used for subsequent frequency domain filtering processing to eliminate abnormal vascular tension interference.

[0166] Preferably, the compensated blood pressure value output module 650 provided in this application is configured with the following units:

[0167] The frequency domain convolution reconstruction unit is used to perform frequency domain convolution processing on the time domain envelope and the blood vessel volume compensation curve. By multiplying the envelope spectrum with the transfer function of the compensation filter in the frequency domain and performing an inverse Fourier transform to convert it to the time domain, the reconstructed compensation envelope is generated.

[0168] The differential feature extraction unit is used to perform differential feature extraction processing on the reconstructed compensation envelope. It finds the maximum point by calculating the first derivative and detects the position point where the second derivative first turns from negative to positive, generating systolic blood pressure feature time and diastolic blood pressure feature time.

[0169] The compensated blood pressure value mapping unit is used to perform pressure value mapping processing on characteristic times of systolic blood pressure and characteristic times of diastolic blood pressure. Based on the characteristic times, it queries the cuff pressure value at the corresponding time and generates a compensated blood pressure value that includes compensated systolic blood pressure value and compensated diastolic blood pressure value.

[0170] Preferably, the differential feature extraction unit includes a first-order derivative sequence generation subunit, a systolic blood pressure feature time detection subunit, and a diastolic blood pressure feature time detection subunit. Specifically, the first-order derivative sequence generation subunit performs first-order derivative calculation on the reconstructed compensated envelope, calculating the instantaneous slope value of each sampling point on the envelope using the central difference method to generate a first-order derivative sequence. The systolic blood pressure feature time detection subunit performs extreme point detection on the first-order derivative sequence, scanning the entire sequence to find the maximum positive value point and recording its corresponding time to generate the systolic blood pressure feature time. The diastolic blood pressure feature time detection subunit performs second-order derivative calculation on the reconstructed compensated envelope, calculating the curvature change value of each sampling point on the envelope using the second-order central difference method, and searching for the first point where the curvature changes from negative to positive after the systolic blood pressure feature time to generate the diastolic blood pressure feature time.

[0171] In one embodiment, this application also provides a smartwatch, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the above-described method for measuring blood pressure compensation in motion scenarios.

[0172] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of those different embodiments or examples.

[0173] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The components described as separate parts may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this disclosure according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0174] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various variations or substitutions within the technical scope disclosed in this application, and these should all be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for blood pressure compensation measurement in a sports setting, characterized in that, Includes the following steps: S1: Process the motion acceleration signal collected by the triaxial accelerometer and the heart rate signal collected by the heart rate sensor, calculate the motion intensity index based on acceleration and the heart rate recovery slope based on the heart rate change rate, and generate a lactate accumulation indicator based on the preset intensity threshold and the preset recovery threshold. S2: Process the cuff pressure oscillation wave signal collected by the pressure sensor, extract the intrinsic mode functions in the 0.5-5Hz frequency band and perform Hilbert transform to generate the time domain envelope; S3: Process the time-domain envelope, calculate the change in the width of the main peak of the envelope relative to the static reference and the energy ratio of the 8-12Hz frequency band to the 1-4Hz frequency band, and generate the tension abnormality degree characterizing vascular tension abnormality. S4: Process the tension abnormality, calculate the vasodilation amount obtained through the nonlinear mapping function, and construct a vascular volume compensation curve by combining it with the exponential decay time function. S5: Process the time-domain envelope and the vascular volume compensation curve, reconstruct the compensated envelope through frequency-domain convolution operation, locate the maximum slope point and the second-order positive change point on the reconstructed envelope, and output the compensated blood pressure value used to eliminate motion interference.

2. The method according to claim 1, characterized in that, S1 includes: S11: Perform cube root integration on the motion acceleration signal acquired by the triaxial accelerometer, calculate the integral value of the cube root of the triaxial composite acceleration within the time window, and generate the motion intensity index. S12: Perform linear regression processing on the heart rate signal collected by the heart rate sensor, calculate the slope value of the heart rate decrease rate after exercise stops, and generate the heart rate recovery slope. S13: Perform dual threshold logic judgment on the exercise intensity index and the heart rate recovery slope. When the exercise intensity exceeds the preset intensity threshold and the heart rate recovery rate is lower than the preset recovery threshold, generate an active lactic acid accumulation flag.

3. The method according to claim 1, characterized in that, S2 includes: S21: Perform empirical mode decomposition on the cuff pressure oscillation wave signal acquired by the pressure sensor, decompose the signal into multiple intrinsic mode function components, and extract the intrinsic mode functions in the 0.5-5Hz frequency band. S22: Perform Hilbert transform on the intrinsic mode functions, calculate the instantaneous amplitude of the analytic signal, and generate the initial envelope; S23: The initial envelope is smoothed and filtered, and high-frequency fluctuation interference is eliminated by calculating the mean within the time window to generate an optimized time-domain envelope.

4. The method according to claim 1, characterized in that, S3 includes: S31: Perform main peak detection processing on the time domain envelope, locate the global maximum point of the envelope and measure its half-width, and generate the main peak width change. S32: Perform frequency domain energy analysis on the time domain envelope, calculate the energy distribution ratio of the 8-12Hz band to the 1-4Hz band, and generate the frequency domain characteristic ratio. S33: The change in the main peak width and the ratio of the frequency domain features are weighted and fused, and the comprehensive offset is calculated in combination with the static reference value to generate the tension anomaly degree. The tension anomaly degree is used to indicate the degree of vascular tension abnormality caused by lactic acid accumulation.

5. The method according to claim 4, characterized in that, The formula for calculating the tension anomaly is: ； in, The degree of vascular tension abnormality is characterized by the extent of abnormal vascular tension. This refers to the change in the width of the main peak. The static reference half-height width, The ratio of frequency domain features. For time-domain feature weighting factors, This is the frequency domain feature weighting factor.

6. The method according to claim 1, characterized in that, S4 includes: S41: Perform nonlinear function mapping on the tension abnormality, and convert the tension abnormality into the proportion of vascular volume change through hyperbolic tangent transformation to generate vascular diastole. S42: The vasodilation volume is modeled by exponential decay, and the time decay function is applied to simulate the dynamic recovery process of vascular volume during lactic acid metabolism, and a volume change curve that decreases with time is constructed. S43: Perform baseline superposition processing on the volume change curve, add the volume change curve to the preset static vascular volume reference value to generate a vascular volume compensation curve, which is used for subsequent frequency domain filtering processing to eliminate abnormal vascular tension interference.

7. The method according to any one of claims 1-6, characterized in that, S5 includes: S51: Perform frequency domain convolution processing on the time domain envelope and the blood vessel volume compensation curve. By multiplying the envelope spectrum with the transfer function of the compensation filter in the frequency domain and performing inverse Fourier transform to convert it to the time domain, a reconstructed compensation envelope is generated. S52: Perform differential feature extraction processing on the reconstructed compensation envelope, find the maximum value point by calculating the first derivative and detect the position point where the second derivative first turns from negative to positive, and generate the characteristic time of systolic blood pressure and diastolic blood pressure. S53: Perform pressure value mapping processing on the characteristic moments of systolic blood pressure and diastolic blood pressure, query the cuff pressure value at the corresponding moment according to the characteristic moment, and generate a compensated blood pressure value that includes compensated systolic blood pressure value and compensated diastolic blood pressure value.

8. The method according to claim 7, characterized in that, S52 includes: S521: Perform first-order derivative calculation on the reconstructed compensation envelope, calculate the instantaneous slope value of each sampling point on the envelope using the central difference method, and generate a first-order derivative sequence; S522: Perform extreme point detection processing on the first derivative sequence, scan the entire sequence to find the maximum positive value point and record its corresponding time, and generate the systolic blood pressure characteristic time. S523: Perform second derivative calculation on the reconstructed compensation envelope, calculate the curvature change value of each sampling point on the envelope using the second-order central difference method, and search for the first point where the curvature changes from negative to positive after the systolic pressure characteristic moment to generate the diastolic pressure characteristic moment.

9. A blood pressure compensation measurement device for sports scenarios, characterized in that, The device includes: The lactate accumulation marker generation module is used to process the motion acceleration signal collected by the triaxial accelerometer and the heart rate signal collected by the heart rate sensor, calculate the exercise intensity index based on acceleration and the heart rate recovery slope based on the heart rate change rate, and generate a lactate accumulation marker based on a preset intensity threshold and a preset recovery threshold. The time-domain envelope generation module is used to process the cuff pressure oscillation wave signal collected by the pressure sensor, extract the intrinsic mode functions in the 0.5-5Hz frequency band and perform Hilbert transform to generate the time-domain envelope. The tension anomaly calculation module is used to process the time domain envelope, calculate the change in the width of the main peak of the envelope relative to the static reference and the energy ratio of the 8-12Hz frequency band to the 1-4Hz frequency band, and generate a tension anomaly characterizing vascular tension abnormality. The vascular volume compensation curve construction module is used to process the tension abnormality, calculate the vascular dilation amount obtained through a nonlinear mapping function, and construct a vascular volume compensation curve by combining it with an exponential decay time function. The compensated blood pressure output module is used to process the time-domain envelope and the vascular volume compensation curve, reconstruct the compensated envelope through frequency-domain convolution operation, locate the maximum slope point and the second-order positive change point on the reconstructed envelope, and output the compensated blood pressure value to eliminate motion interference.

10. A smartwatch, comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the method of any one of claims 1 to 8.

Images