A signal anti-interference processing method and system for assisting a smart watch in measuring blood pressure

Abstract

This invention provides a signal anti-interference processing method and system for assisting smartwatches in measuring blood pressure, relating to the field of medical monitoring signal processing technology. First, it collects the initial pulse wave signal and motion state data of the target user. The current interference level is quantified by calculating a motion interference intensity coefficient. When this coefficient is within a preset range, a dynamically adapted multi-layer filtering fusion is performed based on the initial pulse wave signal and the interference intensity coefficient. Then, variational mode decomposition is performed on the fused filtered signal to remove respiratory fluctuation interference. Finally, the blood pressure measurement result is obtained from the pure target pulse wave signal. This solves the technical problem of insufficient blood pressure measurement accuracy caused by motion artifacts, respiratory fluctuations, and other multi-source interference leading to pulse wave signal distortion when smartwatches use the oscilloscope method to measure wrist blood pressure. It has the technical effect of accurately processing multi-source interference, improving the purity of the pulse wave signal, and reducing blood pressure measurement errors.

Description

Technical Field

[0001] This invention relates to the field of medical monitoring signal processing technology, and in particular to a signal anti-interference processing method and system for assisting smartwatches in measuring blood pressure. Background Technology

[0002] Hypertension, a prevalent chronic cardiovascular disease worldwide, has an increasing incidence rate. Regular and accurate blood pressure monitoring has become a core component of disease prevention, diagnosis, and treatment management, directly impacting patient health risk control and treatment optimization. With the rapid development of wearable medical technology, smart blood pressure monitors, leveraging their portability and real-time capabilities, are gradually replacing traditional desktop blood pressure monitors and becoming an essential tool for daily blood pressure monitoring. Their core technology utilizes the medically recognized oscillometric measurement principle, collecting blood pressure data through a close fit to the wrist. This meets users' health monitoring needs in various scenarios, including home and outdoor settings, and drives the development of non-invasive blood pressure monitoring technology towards lightweight and convenient designs.

[0003] However, the wrist, as a frequently used part of the human body, presents unique challenges to oscillometric blood pressure measurement due to its complex wearing environment. On the one hand, arm movements, muscle contractions, and slight vibrations in the external environment during daily activities can easily generate motion artifacts during measurement, which are directly superimposed on the pulse wave amplitude signal. On the other hand, the inherent characteristics of human respiratory fluctuations and heart rate can easily lead to signal aliasing, further masking the effective characteristics of the pulse wave. Although existing smartwatches have attempted to apply oscillometric methods to wrist blood pressure measurement, their practical application effectiveness remains significantly limited due to limitations in processing multi-source interference from the wrist.

[0004] The core flaw in existing technologies lies in the lack of specificity and synergy in interference handling solutions for wrist blood pressure measurement scenarios. Most products only employ a single filtering method (such as low-pass filtering) to process interference, which cannot effectively distinguish the different interference characteristics of motion artifacts and respiratory fluctuations. Furthermore, it is difficult to suppress interference while ensuring that the effective pulse wave signal does not attenuate. For example, while low-pass filtering can filter high-frequency motion interference to some extent, it easily leads to the loss of high-frequency characteristics of the pulse wave. Some algorithms designed for single types of interference cannot cope with complex interference scenarios involving both motion and respiration. This directly results in high noise levels in the pulse wave signals collected by existing smartwatches, leading to significant errors in subsequent blood pressure calculations. This makes it difficult to achieve medical-grade accuracy, providing users with reliable blood pressure monitoring data and limiting the application of smartwatches in professional health management. There is an urgent need for a technical approach that can adapt to complex wrist scenarios and solve the problem of multi-source interference to improve the accuracy and reliability of wrist blood pressure measurement. Summary of the Invention

[0005] To address the aforementioned shortcomings in existing technologies, the present invention aims to provide a signal anti-interference processing method for assisting smartwatches in measuring blood pressure. This method features the ability to accurately process multi-source interference, improve the purity of pulse wave signals, and reduce blood pressure measurement errors.

[0006] The above-mentioned objective of this invention is achieved through the following technical solution: A signal anti-interference processing method for assisting smartwatches in measuring blood pressure includes: In response to a blood pressure measurement command, the system acquires the initial pulse wave signal of the target user. Acquire the motion state data of the target user, and use the motion state data to calculate the motion interference intensity coefficient; When the motion interference intensity coefficient is within a preset interference intensity range, multi-layer filtering and fusion are performed based on the initial pulse wave signal and the motion interference intensity coefficient to obtain a fused filtered signal. The target pulse wave signal is obtained by performing variational mode decomposition on the fused filtered signal. The target pulse wave signal is used to convert blood pressure to obtain the blood pressure measurement result of the target user.

[0007] By adopting the above technical solution and using a full-process design of "motion interference quantification - multi-layer filtering anti-interference - mode decomposition denoising - accurate blood pressure conversion", the system achieves layered suppression of multi-source noise such as motion artifacts and respiratory interference in blood pressure measurement of smartwatches, effectively improving the purity of pulse wave signals in dynamic scenarios, thereby significantly improving the accuracy and reliability of blood pressure measurement.

[0008] Preferably, the step of acquiring the motion state data of the target user and calculating the motion interference intensity coefficient using the motion state data includes: Acquire the motion state data of the target user, wherein the motion state data is the change in three-axis acceleration; Calculate the Euclidean norm of the triaxial acceleration change corresponding to the aforementioned triaxial acceleration change; The motion disturbance intensity coefficient is obtained by calculating the ratio between the Euclidean norm of the triaxial acceleration change and the preset maximum disturbance acceleration.

[0009] By adopting the above technical solution, the motion amplitude is quantified using the Euclidean norm of the triaxial acceleration change, and then standardized by the ratio with the preset maximum interference acceleration. This allows for precise quantification of the degree of interference of wrist movement on the pulse wave signal, providing an objective and continuous basis for the interference intensity in subsequent adaptive anti-interference processing, and avoiding the problems of over-filtering or under-filtering.

[0010] Preferably, when the motion interference intensity coefficient is within a preset interference intensity range, multi-layer filtering and fusion are performed based on the initial pulse wave signal and the motion interference intensity coefficient to obtain a fused filtered signal, including: When the motion interference intensity coefficient is within the preset interference intensity range, the initial pulse wave signal is used to perform wavelet transform to obtain the first filtered signal. The motion interference intensity coefficient and the initial pulse wave signal are used to perform Kalman adaptive filtering to obtain the second filtered signal; The first filtered signal and the second filtered signal are weighted together with the motion interference intensity coefficient to obtain the fused filtered signal.

[0011] By adopting the above technical solution, a dual-path design is used to preserve signal details with wavelet transform and suppress motion artifacts with Kalman adaptive filtering. Combined with dynamic weighted fusion of motion interference intensity coefficients, the subtle features of the pulse wave are preferentially preserved in weak interference scenarios, while motion artifact suppression is enhanced in strong interference scenarios. This achieves an adaptive balance between anti-interference capability and signal fidelity under different interference intensities.

[0012] Preferably, the step of performing Kalman adaptive filtering using the motion interference intensity coefficient and the initial pulse wave signal to obtain the second filtered signal includes: Determine the Kalman filter parameters, which include the state vector, state transition matrix, observation matrix, observation noise variance, and basic process noise matrix, and initialize the state vector and covariance matrix of the previous time step. The basic process noise matrix is ​​updated using the motion disturbance intensity coefficient to obtain an adaptive process noise matrix; The predicted state vector is obtained by multiplying the state transition matrix with the state vector of the previous time step. The prediction covariance matrix is ​​determined using the state transition matrix, the covariance matrix of the previous time step, the transpose of the state transition matrix, and the adaptive process noise matrix. The Kalman gain is calculated using the prediction covariance matrix, the observation matrix, and the observation noise variance. The state vector at the current moment is obtained by updating the state vector using the Kalman gain, the observation matrix, and the initial pulse wave signal. The first component is extracted from the state vector at the current moment as the second filtered signal.

[0013] By adopting the above technical solution and employing an adaptive strategy that dynamically adjusts the process noise matrix, combined with a recursive process of state prediction, covariance update, and Kalman gain optimization, the dynamic changes of the pulse wave signal can be accurately tracked, motion artifacts of different intensities can be effectively suppressed, and the real-time performance and stability of the signal can be guaranteed, providing a high-quality anti-interference signal for subsequent fusion filtering.

[0014] Preferably, the step of performing variational mode decomposition on the fused filtered signal to obtain the target pulse wave signal includes: Determine the variational mode decomposition parameters; The variational mode decomposition parameters and the fused filtered signal input are used to perform decomposition using a preset variational mode decomposition iterative formula to obtain multiple decomposed mode components; Remove decomposition mode components that fall within a preset breathing frequency range from the multiple decomposition mode components; The remaining multiple decomposed mode components are reconstructed to obtain the target pulse wave signal.

[0015] By adopting the above technical solution and using a directional decomposition strategy with preset respiratory rate constraints, respiratory interference is stripped to a specific modal component and eliminated. The pure pulse wave signal is then obtained by reconstructing the remaining components, which achieves precise directional suppression of respiratory interference and further purifies the pulse wave signal, providing a high-fidelity input for blood pressure conversion.

[0016] Preferably, the step of converting the target pulse wave signal into blood pressure to obtain the blood pressure measurement result of the target user includes: Extract the characteristic envelope of amplitude variation with wrist airbag pressure from the target pulse wave signal; Based on preset systolic and diastolic blood pressure thresholds, corresponding airbag pressure points are matched from the feature envelope and used as systolic and diastolic blood pressure measurements, respectively. By integrating the systolic blood pressure measurement value and the diastolic blood pressure measurement value, the blood pressure measurement result of the target user is obtained.

[0017] By adopting the above technical solution, the feature envelope is extracted based on the oscillometric method. Combined with clinically calibrated proportional thresholds to match systolic and diastolic blood pressure, blood pressure characteristics can be accurately captured from pure target pulse wave signals. This effectively avoids the interference of noise on envelope extraction and significantly improves the clinical reliability of systolic and diastolic blood pressure measurements.

[0018] Preferred options also include: When the motion interference intensity coefficient is not within the preset interference intensity range, a blood pressure measurement pause command is generated and fed back to the target user.

[0019] By adopting the above technical solution, when the intensity of motion interference exceeds the processing range of the algorithm, a pause measurement command is generated in a timely manner and fed back to the user. This can avoid the output of invalid measurement results, reduce user misunderstanding, improve the intelligence of the measurement process and user experience, and ensure the reliability of blood pressure measurement results.

[0020] The second objective of this invention is to provide a signal anti-interference processing system for assisting smartwatches in measuring blood pressure, which features the ability to accurately process multi-source interference, improve the purity of pulse wave signals, and reduce blood pressure measurement errors.

[0021] The second objective of this invention is achieved through the following technical solution: A signal anti-interference processing system for assisting smartwatches in measuring blood pressure includes: The instruction response module is used to respond to received blood pressure measurement instructions and acquire the initial pulse wave signal of the target user. The data processing module is used to acquire the motion state data of the target user and use the motion state data to calculate the motion interference intensity coefficient. The filtering and fusion module is used to perform multi-layer filtering and fusion based on the initial pulse wave signal and the motion interference intensity coefficient when the motion interference intensity coefficient is within a preset interference intensity range, so as to obtain a fused filtered signal. The mode decomposition module is used to perform variational mode decomposition on the fused filtered signal to obtain the target pulse wave signal; The measurement output module is used to convert the target pulse wave signal into blood pressure and obtain the blood pressure measurement result of the target user.

[0022] By adopting the above technical solution, and employing a collaborative architecture of five modules—command response, data processing, filtering and fusion, mode decomposition, and measurement output—the command response module quickly responds to blood pressure measurement commands and accurately acquires the initial pulse wave signal, providing a basic input for subsequent processing. The data processing module quantifies the motion interference intensity coefficient, providing an objective and reliable basis for interference judgment in anti-interference processing. The filtering and fusion module achieves multi-layer adaptive filtering and fusion for motion interference of different intensities, balancing anti-interference capability and signal fidelity. The mode decomposition module further removes residual noise such as respiratory interference, purifying the pulse wave signal. The measurement output module completes blood pressure conversion based on the pure target pulse wave signal. Each module performs its function seamlessly, achieving accurate processing of multi-source noise such as motion artifacts and respiratory interference in smartwatch blood pressure measurement. This effectively improves the purity of the pulse wave signal, reduces the impact of interference factors on blood pressure measurement, and ultimately reduces blood pressure measurement error. This perfectly aligns with the second objective of this invention, enabling stable and accurate output of blood pressure measurement results, meeting the practicality and reliability requirements of blood pressure measurement in dynamic scenarios using smartwatches.

[0023] The third objective of this invention is to provide an electronic device that can accurately process multi-source interference, improve the purity of pulse wave signals, and reduce blood pressure measurement errors.

[0024] The above-mentioned objective three of this invention is achieved through the following technical solution: An electronic device includes a memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and execute the signal anti-interference processing method for assisting a smartwatch in measuring blood pressure as described above.

[0025] The fourth objective of this invention is to provide a computer-readable storage medium capable of storing corresponding programs, which facilitates the accurate processing of multi-source interference, improves the purity of pulse wave signals, and reduces blood pressure measurement errors.

[0026] The fourth objective of this invention is achieved through the following technical solution: A computer-readable storage medium storing a computer program capable of being loaded by a processor and executing the signal anti-interference processing method for assisting smartwatch blood pressure measurement as described above.

[0027] In summary, the present invention has at least one of the following beneficial technical effects: 1. This invention provides a signal anti-interference processing method and system for assisting smartwatches in measuring blood pressure. First, it collects the initial pulse wave signal and motion state data of the target user. The current interference level is quantified by calculating a motion interference intensity coefficient. When this coefficient is within a preset range, a dynamically adapted multi-layer filtering fusion is performed based on the initial pulse wave signal and the interference intensity coefficient. Then, variational mode decomposition is performed on the fused filtered signal to remove respiratory fluctuation interference. Finally, the blood pressure measurement result is obtained from the pure target pulse wave signal. This invention achieves dynamic adjustment of the anti-interference strategy through the quantification of the motion interference intensity coefficient, and the multi-layer filtering fusion can target motion artifacts of different intensities. Adaptive suppression avoids the shortcomings of single filtering algorithms, such as over-smoothing or insufficient suppression, when dealing with complex motion interference. Variational mode decomposition can target and precisely isolate interference components within the respiratory frequency range, solving the signal distortion problem caused by the aliasing of respiratory fluctuations and heart rate. The synergistic effect of the two can significantly improve the purity of the pulse wave signal, thereby greatly reducing blood pressure measurement errors. At the same time, the entire algorithm process is adapted to the hardware computing power of smartwatches, ensuring both anti-interference effect and real-time measurement. Ultimately, it enables smartwatches to achieve medical-grade accuracy in wrist blood pressure measurement, effectively solving the pain point of insufficient measurement accuracy caused by multi-source interference in existing technologies.

[0028] 2. The motion interference intensity coefficient calculated from motion state data can accurately quantify the strength of current motion artifacts. The larger the coefficient, the more intense the motion and the stronger the artifact interference; the smaller the coefficient, the milder the motion and the weaker the artifact interference. This coefficient serves as the core adjustment basis for multi-layer filtering fusion, enabling the fusion process to dynamically adjust the filtering strategy and parameter ratio according to the interference intensity. When the motion interference intensity coefficient is high (strong motion artifact scenario), multi-layer filtering fusion will focus on strengthening the suppression of high-frequency and large-amplitude motion artifacts, prioritizing the thoroughness of interference stripping; when the motion interference intensity coefficient is low (weak motion artifact scenario), multi-layer filtering fusion will reduce the suppression intensity, focusing on preserving the effective details of the pulse wave and avoiding signal distortion caused by over-filtering. It is through the dynamic control of the multi-layer filtering fusion process by the motion interference intensity coefficient that differentiated and adaptive suppression of motion artifacts of different intensities is achieved. This solves the problem of insufficient suppression under strong interference and avoids the defect of over-filtering under weak interference, ultimately accurately adapting to the variable motion artifact scenarios in wrist measurement. Attached Figure Description

[0029] Figure 1 This is a flowchart illustrating the steps of a signal anti-interference processing method for assisting smartwatches in measuring blood pressure, provided in an embodiment of the present invention.

[0030] Figure 2 This is a flowchart illustrating the basic structure of a smartwatch for blood pressure measurement provided in an embodiment of the present invention.

[0031] Figure 3 This is a flowchart of the anti-interference algorithm module provided in the embodiment of the present invention.

[0032] Figure 4 This is a structural block diagram of a signal anti-interference processing system for assisting smartwatches in measuring blood pressure, provided in an embodiment of the present invention. Detailed Implementation

[0033] This invention provides a signal anti-interference processing method and system for assisting smartwatches in measuring blood pressure. It addresses the technical problem of insufficient blood pressure measurement accuracy caused by motion artifacts, respiratory fluctuations, and other multi-source interference leading to pulse wave signal distortion when smartwatches measure wrist blood pressure using the oscillometric method. The method effectively processes multi-source interference, improves pulse wave signal purity, and reduces blood pressure measurement errors.

[0034] To make the objectives, features, and advantages of this invention more apparent and understandable, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described below are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0035] It should be noted that, in the embodiments of this invention, when the relevant object information and other related data are used in specific products or technologies, permission or consent from the object is required, and the collection, use, and processing of the relevant data must comply with the relevant laws, regulations, and standards of the relevant countries and regions. In other words, if the embodiments of this invention involve data related to an object, it must be obtained with the object's authorization and consent, the authorization and consent of relevant departments, and in accordance with the relevant laws, regulations, and standards of the country and region. If personal information is involved in the embodiments, the acquisition of all personal information requires the individual's consent; if sensitive information is involved, the separate consent of the information subject is required. The embodiments also need to be implemented with the object's authorization and consent.

[0036] It should be noted that the terms "first," "second," etc., used in this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this disclosure described herein can be implemented in orders other than those illustrated or described herein. The implementations described in the following exemplary embodiments do not represent all implementations consistent with this disclosure.

[0037] Furthermore, the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article, unless otherwise specified, generally indicates that the preceding and following related objects have an "or" relationship.

[0038] This invention aims to solve the technical problem of signal distortion caused by multiple sources of interference, such as motion artifacts and respiratory fluctuations, in wrist blood pressure measurement. Based on the oscillometric measurement principle, this method employs a three-level processing mechanism of "motion state recognition - hierarchical filtering - interference removal," combined with hardware sensor data linkage. First, an accelerometer is used to determine the interference intensity. Then, wavelet transform and adaptive intelligent filtering are used to remove motion and vibration noise. Finally, variational mode decomposition is used to remove respiratory interference, achieving efficient extraction of a pure pulse wave signal. A smartwatch using this method utilizes a micro-pump and flexible airbag inflation / deflation mechanism. A high-precision pressure sensor collects the pulse wave amplitude signal, which, after anti-interference processing, is adapted to the oscillometric calculation logic. This ensures signal stability even in complex wearing scenarios, providing a reliable data foundation for accurate blood pressure calculation and is suitable for medical-grade wrist blood pressure measurement devices. Example

[0039] Please see Figure 1 The present invention provides a signal anti-interference processing method for assisting smartwatches in measuring blood pressure, comprising: The signal anti-interference processing method for blood pressure measurement using smartwatches is a signal processing approach specifically designed for smartwatch blood pressure measurement scenarios. It improves the quality of the pulse wave signal by suppressing noise such as motion artifacts and respiratory interference, thereby enhancing the accuracy and reliability of blood pressure measurement. The blood pressure measurement command is the user-initiated command through the smartwatch's interface to trigger the blood pressure measurement process. The initial pulse wave signal is the raw pressure signal directly collected by the smartwatch's pressure sensor during the inflation and deflation of the wrist airbag, containing components such as the effective pulse wave, motion artifacts, and respiratory interference. Motion state data is data collected by the smartwatch's built-in accelerometer, reflecting the amplitude of the user's wrist movements. The motion interference intensity coefficient is a standardized coefficient quantifying the degree of interference of wrist movements on the pulse wave signal, ranging from 0 to 1, with higher values ​​indicating stronger motion interference. The preset interference intensity range is the range of motion interference intensity that the anti-interference algorithm can effectively handle, calibrated through prior experiments, typically including static, slight motion, and weak motion levels. Multi-layer filtering fusion is a multi-path filtering strategy that combines wavelet transform and Kalman adaptive filtering. By dynamically weighting and fusing the two filtering results, it achieves adaptive anti-interference under different interference intensities. The fused filtered signal is a pulse wave signal that combines anti-interference capability and signal fidelity after wavelet transform and Kalman adaptive filtering, and then dynamically weighted and fused. Variational mode decomposition is an adaptive signal decomposition algorithm that can decompose complex signals into multiple mode components with specific center frequencies, used to selectively remove noise such as respiratory interference. The target pulse wave signal is a pure pulse wave signal that has been processed by multi-layer filtering fusion and variational mode decomposition, removing motion artifacts, respiratory interference and other multi-source noise. It is the core input for blood pressure conversion. Blood pressure conversion is based on the oscillometric principle, extracting the feature envelope from the pure pulse wave signal and matching it to obtain systolic and diastolic blood pressure. The blood pressure measurement result is the final measurement output containing systolic and diastolic blood pressure values, which can be directly displayed on the smartwatch screen for users to view.

[0040] Step 101: Respond to the blood pressure measurement command and obtain the initial pulse wave signal of the target user.

[0041] In this embodiment of the invention, when the smartwatch receives a blood pressure measurement command triggered by the target user, it immediately activates the wrist signal acquisition module to acquire the initial pulse wave signal corresponding to the wrist artery of the target user. This initial pulse wave signal serves as the basic data source for subsequent anti-interference processing and blood pressure conversion.

[0042] Step 102: Obtain the motion state data of the target user and use the motion state data to calculate the motion interference intensity coefficient.

[0043] In this embodiment of the invention, considering that the user's motion state during wrist measurement will generate motion artifacts that interfere with the purity of the initial pulse wave signal, the motion state acquisition module is simultaneously activated after acquiring the target user's initial pulse wave signal to obtain the motion state data of the target user during the current measurement process. This motion state data can intuitively reflect the amplitude of the user's wrist movement and vibration. Based on this motion state data, a quantification calculation is performed to obtain the motion interference intensity coefficient. This coefficient is used to accurately characterize the degree of interference of motion artifacts on the initial pulse wave signal in the current measurement scenario. Subsequently, based on this coefficient, it will be determined whether to enter the multi-layer filtering fusion process to adapt to the signal processing requirements under different interference intensities.

[0044] Furthermore, step 102 may include the following sub-steps: S11. Obtain the motion state data of the target user, which is the change in three-axis acceleration. S12. Calculate the Euclidean norm of the triaxial acceleration change corresponding to the triaxial acceleration change. S13. The motion disturbance intensity coefficient is obtained by calculating the ratio between the Euclidean norm of the triaxial acceleration change and the preset maximum disturbance acceleration.

[0045] The three-axis acceleration change is the instantaneous change value of the wrist movement acceleration fluctuation collected by the smartwatch's accelerometer in three directions; the Euclidean norm of the three-axis acceleration change is a scalar value obtained by calculating the vector magnitude of the three-axis acceleration change, used to integrate motion information in three directions; the preset maximum interference acceleration is the maximum motion acceleration threshold that the smartwatch's anti-interference algorithm can handle, calibrated in previous experiments, and is usually set with reference to gravitational acceleration.

[0046] In this embodiment of the invention, to accurately quantify the interference of motion artifacts on the initial pulse wave signal, and combining the acquisition capability of the smartwatch's built-in accelerometer, step S11 is executed to obtain the triaxial acceleration changes of the wrist during the target user's measurement process. This data corresponds to the acceleration fluctuations of the wrist in the X, Y, and Z directions, and can comprehensively reflect the amplitude of wrist movement and vibration. Subsequently, step S12 is executed to calculate the Euclidean norm of the triaxial acceleration changes, integrating the motion information in the three directions to avoid the limitations of acceleration data in a single direction. The mathematical formula is as follows:

[0047] in The Euclidean norm represents the change in triaxial acceleration. These represent the changes in triaxial acceleration in the X, Y, and Z directions, respectively. Finally, step S13 is executed to calculate the ratio between the Euclidean norm and the preset maximum disturbance acceleration, thereby standardizing and quantifying the disturbance intensity to obtain the motion disturbance intensity coefficient. The mathematical formula is as follows:

[0048] in The motion interference intensity coefficient, To preset the maximum interference acceleration (calibrated in previous experiments, corresponding to the maximum acceleration threshold that the smartwatch can eliminate through filtering), this sub-step process can accurately obtain the coefficient characterizing the intensity of the current motion interference, providing a reliable basis for the interference interval judgment and multi-layer filtering fusion parameter adaptation in step 103.

[0049] Step 103: When the motion interference intensity coefficient is within the preset interference intensity range, multi-layer filtering and fusion are performed based on the initial pulse wave signal and the motion interference intensity coefficient to obtain the fused filtered signal.

[0050] In this embodiment of the invention, the calculated motion interference intensity coefficient is first judged within a certain range. The preset interference intensity range corresponds to a reasonable range in the smartwatch wrist blood pressure measurement where interference can be eliminated through filtering. When the coefficient is within this range, it indicates that the current motion artifact interference can be suppressed through signal processing. At this time, the initial pulse wave signal obtained in step 101 and the motion interference intensity coefficient calculated in step 102 are combined to perform multi-layer filtering fusion processing. The motion interference intensity coefficient is used to adapt the filtering fusion parameters to suppress the motion artifacts corresponding to the current interference intensity in a targeted manner, while retaining the effective features in the initial pulse wave signal, and finally obtaining a preliminarily purified fused filtering signal.

[0051] Furthermore, step 103 may include the following sub-steps: S21. When the motion interference intensity coefficient is within the preset interference intensity range, the initial pulse wave signal is used to perform wavelet transform to obtain the first filtered signal. S22. Kalman adaptive filtering is performed using the motion interference intensity coefficient and the initial pulse wave signal to obtain the second filtered signal; S23. The first and second filtered signals are weighted together with the motion interference intensity coefficient to obtain the fused filtered signal.

[0052] Wavelet transform is an algorithm that uses the Dobessi 4 wavelet basis function to decompose signals at multiple scales. It can separate signals into approximate components and detail components, which are used for preliminary denoising and preservation of signal details. The first filtered signal is a filtered signal reconstructed after wavelet transform decomposition and adaptive threshold denoising. It can preserve the detailed features of the pulse wave well in weak interference scenarios. Kalman adaptive filtering is a recursive filtering algorithm based on a state-space model. It can adapt to different intensities of motion interference by dynamically adjusting the process noise matrix, thereby suppressing motion artifacts. The second filtered signal is a filtered signal obtained after Kalman adaptive filtering. It has a strong anti-motion artifact capability in strong interference scenarios. Weighting is based on the dynamic allocation of weights between the first and second filtered signals according to the motion interference intensity coefficient to achieve adaptive fusion under different interference intensities.

[0053] In this embodiment of the invention, when the motion interference intensity coefficient is within a preset interference intensity range, the preset interference intensity range is divided into a stationary level (…). ≤0.1), light exercise level (0.1< ≤0.3), weak motion level (0.3< ≤0.5), first execute S21, selecting the Daubechies4 (db4) wavelet basis function. The Dobessi 4 wavelet basis function is used to perform a 5-level multi-scale decomposition on the initial pulse wave signal, decomposing the signal into an approximate component A5 (0-1.953125Hz, corresponding to the low-frequency effective pulse wave signal) and detail components D1-D5. The frequency ranges of each detail component are: D1: 31.25-62.5Hz (high-frequency noise), D2: 15.625-31.25Hz, D3: 7.8125-15.625Hz, D4: 3.90625-7.8125Hz, D5: 1.953125-3.90625Hz (main range of motion artifacts). Subsequently, differentiated denoising adaptation is performed for different levels of interference: when in the static level, the threshold setting is more lenient to preserve the subtle features of the pulse wave to the greatest extent; when in the slight motion level, a standard adaptive thresholding method is used. The threshold calculation formula is as follows:

[0054] Here, the Adaptive Denoising Threshold is a core quantization parameter in the wavelet transform denoising process, used to accurately distinguish noise figures from effective signal figures in wavelet detail components. denoted as the noise standard deviation of the detail components, and N as the signal length.

[0055] In the detail components of wavelet decomposition, the parts with absolute values ​​less than 0 are judged as noise and set to zero, while the parts with absolute values ​​greater than or equal to 0 are retained. This achieves the denoising effect of removing high-frequency noise and retaining the effective high-frequency features of the pulse wave, especially in weak interference scenarios, where it can better balance noise suppression and signal fidelity.

[0056] By using this threshold to remove high-frequency motion / vibration noise from the detail components while retaining the effective high-frequency components of the pulse wave, this branch can effectively handle weak interference scenarios. Significant advantages are observed at levels ≤0.3). When the motion level is weak, the threshold of the D5 component in the main region of motion artifacts is appropriately increased to strengthen the suppression of motion artifacts in this region, avoid residual interference affecting subsequent processing, and finally reconstruct a first filtered signal adapted to the current interference level. Then, S22 is executed, using the motion interference intensity coefficient as the dynamic adjustment basis of the Kalman adaptive filter, and combining it with the initial pulse wave signal to complete the filtering process, obtaining a second filtered signal that is more adapted to the current motion interference intensity. Finally, S23 is executed, and the first filtered signal and the second filtered signal are weighted and fused, with the fusion formula being: ,in To fuse filtered signals, This is the first filtered signal. This is the second filtered signal. The motion interference intensity coefficient is used, and the fusion process follows a weight allocation logic: in weak interference scenarios ( ≤0.3) ≥70%, by Dominant, prioritizing the preservation of detailed features of the pulse wave; in scenarios with strong interference ( ≥0.5), ≥50%, by Dominant, enhancing the suppression of motion artifacts; in moderate interference scenarios (0.3 < Under <0.5), the two branches work together to balance signal fidelity and interference suppression, thereby achieving adaptive fusion under different interference intensities and obtaining a fused filtered signal that has both anti-interference effect and signal fidelity, providing high-quality input for subsequent variational mode decomposition.

[0057] Furthermore, S22 may include the following sub-steps: S221. Determine the Kalman filter parameters, which include the state vector, state transition matrix, observation matrix, observation noise variance, and basic process noise matrix, and initialize the state vector and covariance matrix of the previous time step. S222. The basic process noise matrix is ​​updated using the motion disturbance intensity coefficient to obtain the adaptive process noise matrix; S223. The predicted state vector is obtained by multiplying the state transition matrix with the state vector of the previous time step. S224. Using the state transition matrix, the covariance matrix of the previous time step, the transpose of the state transition matrix, and the adaptive process noise matrix, determine the prediction covariance matrix. S225. Calculate the Kalman gain using the prediction covariance matrix, the observation matrix, and the observation noise variance. S226. Update the state vector using Kalman gain, observation matrix and initial pulse wave signal to obtain the state vector at the current time. S227. Extract the first component from the current state vector as the second filtered signal.

[0058] Kalman filter parameters are the core configuration parameters of Kalman adaptive filtering, including the state vector, state transition matrix, observation matrix, observation noise variance, and basic process noise matrix. The state vector describes the state of the Kalman filter system; in this scheme, it is used to characterize the airbag pressure value and pressure change rate at a given moment. The state transition matrix describes the evolution of the system state from the previous moment to the current moment and is related to the sensor sampling interval. The observation matrix maps the system state vector to the observed values. The observation noise variance quantifies the noise level of the sensor observations. The basic process noise matrix is ​​the initial process noise matrix without motion interference, used to ensure basic measurement accuracy. The state vector from the previous moment is the system state vector from the previous moment, used in the prediction step of the Kalman filter. The covariance matrix from the previous moment is... The system state covariance matrix is ​​used to quantify the uncertainty of state estimation; the adaptive process noise matrix is ​​a process noise matrix dynamically adjusted based on the motion disturbance intensity coefficient. The stronger the disturbance, the larger the matrix value and the stronger the filtering and suppression capability; the predicted state vector is the current state prediction value calculated by the state transition matrix and the previous time step state vector; the prediction covariance matrix is ​​a matrix that quantifies the uncertainty of the predicted state vector, calculated by the state transition matrix, the previous time step covariance matrix, the transpose of the state transition matrix, and the adaptive process noise matrix; the Kalman gain is a coefficient used to balance the weights of the predicted and observed values, calculated by the prediction covariance matrix, the observation matrix, and the observation noise variance; the current time step state vector is the current time step system state vector updated by the Kalman gain, containing more accurate pressure and pressure change rate information.

[0059] In this embodiment of the invention, S221 is first executed to determine the core filtering parameters based on the adaptation requirements of Kalman adaptive filtering: the state vector is defined as:

[0060] in The pressure value at time k (unit: mmHg). Pressure change rate (unit: mmHg / ms); The state transition matrix A is:

[0061] in The sensor sampling interval is fixed at 4ms, corresponding to a 250Hz sampling rate. The observation matrix H is:

[0062] Basic process noise matrix for:

[0063] The observation noise variance R is:

[0064] Initialize the state vector from the previous time step. :

[0065] in This is the initial pressure value; Covariance matrix of the previous time step for:

[0066] Next, step S222 is executed to dynamically adjust the process noise based on the motion disturbance intensity coefficient, thereby obtaining an adaptive process noise matrix. The formula is:

[0067] This design ensures that the stronger the motion interference and the larger the process noise matrix, the stronger the filtering ability to suppress motion interference.

[0068] Then, step S223 is executed, and the predicted state vector is obtained by matrix multiplication of the state transition matrix and the state vector from the previous time step. The formula is:

[0069] Then execute S224 to calculate the prediction covariance matrix. The formula is:

[0070] Among them The covariance matrix of the previous time step is the transpose of the state transition matrix.

[0071] Then, step S225 is executed to calculate the Kalman gain based on the prediction covariance matrix, the observation matrix, and the observation noise variance. The formula is:

[0072] Then execute S226, combining the Kalman gain, the observation matrix, and the initial pulse wave signal (i.e., the sensor's measured pressure value). The state vector is updated using the following formula:

[0073] Finally, execute S227, starting from the current state vector. Extract the first component As the second filtered signal This signal is the initial purification pressure signal after Kalman adaptive filtering, which can effectively suppress motion artifact interference.

[0074] Step 104: Perform variational mode decomposition using the fused filtered signal to obtain the target pulse wave signal.

[0075] In this embodiment of the invention, the fused filtered signal obtained in step 103 has initially suppressed motion artifact interference, but some interference components may still remain. Therefore, the fused filtered signal is used as input to perform variational mode decomposition processing. Through this decomposition method, the residual interference in the signal is further removed, and the effective features of the pulse wave are accurately extracted. Finally, a pure target pulse wave signal is obtained, which provides a reliable data source for subsequent blood pressure conversion and accurate blood pressure measurement results, and achieves thorough layered processing of interference.

[0076] Furthermore, step 104 may include the following sub-steps: S31. Determine the variational mode decomposition parameters; S32. Decompose the mode using variational mode decomposition parameters and fused filtered signal input to a preset variational mode decomposition iterative formula to obtain multiple decomposed mode components; S33. Remove decomposition mode components that fall within the preset respiratory frequency range from multiple decomposition mode components; S34. Reconstruct the remaining multiple decomposed mode components to obtain the target pulse wave signal.

[0077] Variational mode decomposition (VMD) parameters are configuration parameters for VMD, including the number of decomposed modes, respiratory interference mode frequency constraints, and dynamic penalty factors. The preset VMD iterative formula is the core optimization iterative formula for VMD, used to decompose the input signal into multiple modal components, with the goal of minimizing the sum of the bandwidths of each modal component. The decomposed modal components are multiple signal components with specific center frequencies obtained from VMD; this scheme includes respiratory interference-specific components. The preset respiratory frequency range is the human respiratory frequency interval calibrated in previous experiments, used to identify and remove respiratory interference modal components. Reconstruction is the process of superimposing the remaining modal components after removing respiratory interference to obtain a pure target pulse wave signal.

[0078] In this embodiment of the invention, the variational mode decomposition uses the VMD (Variational Mode Decomposition) algorithm. Continuing from the fused filtered signal obtained in the previous step, step S31 is first executed to determine the core parameters of the variational mode decomposition (VMD). The number of decomposition modes is set to n=6, where the 6th mode is a dedicated channel for breathing interference, and its center angular frequency... Constrained by the breathing disturbance mode frequency constraint formula, the formula is:

[0079] in =0.1Hz =0.3Hz, thereby forcing respiratory interference into this dedicated modal channel to achieve targeted stripping; Simultaneously set dynamic penalty factors The formula is:

[0080] in This is the real-time motion interference intensity coefficient (shared with the aforementioned Kalman filter). The stronger the motion interference ( The larger ( The larger the value, the stronger the signal's anti-aliasing capability.

[0081] Next, step S32 is executed, using the fused and filtered signal as the input signal. Based on the above VMD parameters, the preset VMD core objective formula is input for iterative decomposition. The core objective formula is:

[0082] in For sets and conduct Minimize the solution; this is the goal-oriented approach for the entire optimization problem. This is a first-order partial derivative operator with respect to time t, used to calculate the instantaneous frequency correlation characteristics of a signal. For Dirac The function is an impulse function, used here for Hilbert transform preprocessing of the signal to obtain an analytic signal. The kernel function of the Hilbert transform, and Combined with the generating operators that constitute the analytic signal, where The imaginary unit ( ), Let be the time-domain signal of the nth modal component, representing the component of the original signal within a specific frequency band. This is a complex exponential factor used to transform the analytic signal from the time domain to the frequency domain and align it to the center angular frequency of that mode. , "Subject to" is an abbreviation for "constraints". For all modal components The sum equals the input signal (i.e., the fused filtered signal obtained in step 103), ensuring that no signal energy is lost during the decomposition process. After decomposition Each modal component ( ), The center angular frequency of each mode is used as the starting point. The goal is to minimize the sum of the bandwidths of each mode component to achieve accurate signal decomposition, ultimately resulting in six decomposed mode components. .

[0083] Then, step S33 is executed, which identifies and removes the 6th decomposition mode component that falls within the preset respiratory frequency range (0.1Hz~0.3Hz) based on the respiratory interference mode frequency constraint. This component is the respiratory interference component that has been targeted and stripped. Finally, S34 is executed to reconstruct the remaining 5 decomposed mode components, and the output formula is:

[0084] This reconstruction yields the target pulse wave signal stripped of respiratory interference. This provides a high-purity signal input for subsequent blood pressure conversion.

[0085] Step 105: Use the target pulse wave signal to convert blood pressure and obtain the blood pressure measurement result of the target user.

[0086] In this embodiment of the invention, step 104 has obtained a pure target pulse wave signal that has been stripped of motion artifacts, residual interference and other interferences and has completely preserved the effective characteristics of the pulse wave. Based on this reliable target pulse wave signal, blood pressure conversion processing is performed to convert the effective characteristics of the pulse wave into corresponding blood pressure parameters, and then the blood pressure measurement results of the target user are integrated to ensure the accuracy of the blood pressure measurement results. This perfectly echoes the core purpose of the anti-interference processing mentioned above, and finally realizes the accurate measurement of blood pressure on the wrist of the smartwatch, solving the problem of insufficient measurement accuracy caused by multi-source interference in the prior art.

[0087] Furthermore, step 105 may include the following sub-steps: S41. Extract the characteristic envelope of amplitude variation with wrist airbag pressure from the target pulse wave signal; S42. Based on preset systolic and diastolic pressure thresholds, match the corresponding airbag pressure points from the feature envelope, and use them as systolic and diastolic pressure measurements, respectively. S43. Integrate systolic and diastolic blood pressure measurements to obtain the target user's blood pressure measurement results.

[0088] The characteristic envelope is a continuous curve reflecting the change in pulse wave amplitude with wrist cuff pressure, and is the core basis for oscillometric blood pressure measurement. The preset systolic and diastolic blood pressure thresholds are amplitude ratio thresholds based on clinical trials and used to match systolic and diastolic blood pressure. The cuff pressure point is the wrist cuff pressure value on the characteristic envelope corresponding to the amplitude of the systolic and diastolic blood pressure thresholds, which is directly used as the measured values ​​of systolic and diastolic blood pressure. The systolic blood pressure measurement value is the peak pressure in the artery during cardiac contraction, corresponding to the cuff pressure value that matches the amplitude of the preset systolic blood pressure threshold at the rising edge of the characteristic envelope. The diastolic blood pressure measurement value is the trough pressure in the artery during cardiac relaxation, corresponding to the cuff pressure value that matches the amplitude of the preset diastolic blood pressure threshold at the falling edge of the characteristic envelope.

[0089] In this embodiment of the invention, step S41 is first executed. Based on the oscilloscope measurement principle, the peak value of the target pulse wave signal is extracted step by step according to the heartbeat cycle, and the peak amplitude of the pulse wave in each step is extracted. Simultaneously record the wrist airbag pressure value corresponding to the peak amplitude. ,in Given the heartbeat cycle number, the formula for extracting the peak amplitude is:

[0090] in The first The start and end times of the i-th and i-th heartbeat cycles will be obtained beat by beat. and Correlation fitting is performed to form a characteristic envelope in which the amplitude changes continuously with the pressure of the wrist air bladder. This envelope directly reflects the change law of pulse wave amplitude with the rise and fall of air bladder pressure, and is the core basis for blood pressure conversion. Next, execute S42 to first determine the maximum peak amplitude from the characteristic envelope. That is, the amplitude of the vertex of the envelope, and then based on the preset shrinkage pressure ratio coefficient. and diastolic blood pressure proportionality coefficient Calculate the amplitude corresponding to the systolic pressure respectively. for:

[0091] Diastolic pressure corresponding amplitude for:

[0092] in The value is 0.5. The value is 0.8, which represents the standard coefficient for wrist measurements calibrated in previous clinical trials. Subsequently, the amplitude is matched at the rising edge of the characteristic envelope. The point is used as the systolic blood pressure measurement value for the airbag pressure value at that point. At the falling edge of the characteristic envelope, the matching amplitude is equal to The point is used as the diastolic blood pressure measurement value for the airbag pressure value at that point. ; Finally, step S43 is executed to integrate the systolic and diastolic blood pressure measurements to form a complete blood pressure measurement result containing both systolic and diastolic pressure. This result is obtained based on the conversion of a pure target pulse wave signal, effectively avoiding the influence of multiple sources of interference such as motion artifacts and respiratory interference, thus ensuring the accuracy and reliability of wrist blood pressure measurement.

[0093] Furthermore, it may also include the following steps: Step 106: When the motion interference intensity coefficient is not within the preset interference intensity range, a blood pressure measurement pause command is generated and fed back to the target user.

[0094] The blood pressure measurement pause command is a command generated by the system to terminate the current measurement process when the intensity of exercise interference exceeds the preset range, i.e., when the exercise is at the level of strenuous exercise. The feedback is to inform the user of the blood pressure measurement pause command through the smartwatch's display screen, vibration, etc., prompting them to adjust their exercise state and try again.

[0095] In this embodiment of the invention, combining the motion interference intensity coefficient calculated in step 102 and the interference intensity grading rules, when the coefficient is not within the preset interference intensity range (i.e., within the severe motion level), When the value is >0.5 (corresponding to an acceleration change >0.5g), it indicates that the motion artifact interference in the current measurement scenario has exceeded the range that multilayer filtering fusion and variational mode decomposition can handle. If subsequent signal processing and blood pressure conversion are continued, the excessive amount of intense motion interference cannot be removed to obtain a pure pulse wave signal, ultimately leading to excessive deviation in blood pressure measurement results and loss of clinical reference significance. Therefore, a blood pressure measurement pause command will be generated at this time and fed back to the target user, prompting the user to adjust their wrist position and avoid intense motion scenarios. The measurement will continue until the motion interference intensity coefficient falls below the level including the static level (…). ≤0.1), light exercise level (0.1< ≤0.3), weak motion level (0.3< The measurement is restarted only after the preset interference intensity range of ≤0.5 is reached, thus ensuring the effectiveness and accuracy of blood pressure measurement.

[0096] Please participate Figures 2-3 Here is an example of its application: When the target user wears the smart blood pressure watch of this invention and moves slightly in the office (with a slight swing of the arm), the blood pressure measurement function is triggered. The complete execution process is as follows: To initiate a measurement, the user clicks the "Blood Pressure Measurement" button on the watch display, and the watch display sends a blood pressure measurement command to the watch's main control chip.

[0097] The main control chip of the watch controls a micro air pump to inflate the flexible airbag in the wrist. At the same time, a high-precision pressure sensor synchronously collects the raw pressure signal of the airbag, and the watch's accelerometer collects the three-axis acceleration change of the wrist (motion state data). Both types of data are synchronously transmitted to the noise reduction algorithm processing module.

[0098] The interference intensity grading unit within the interference intensity grading and noise reduction algorithm processing module calculates the motion state data: First, calculate the Euclidean norm of the triaxial acceleration changes. Then compared with the preset maximum disturbance acceleration The motion disturbance intensity coefficient is obtained by taking the ratio of 0.5g. =0.2.

[0099] According to the classification rules =0.2 is considered light exercise level (0.1 < If the value is ≤0.3, it is within the preset interference intensity range and there is no need to trigger a pause measurement prompt; instead, it enters the anti-interference filtering process.

[0100] The first-level filtering and noise reduction algorithm processing module starts the first-level filtering subsystem: Wavelet transform unit: The Daubechies4 (db4) wavelet basis function is used to perform a 5-level multi-scale decomposition on the original pressure signal to obtain the approximate component A5 and the detail components D1-D5 (D1: 31.25-62.5Hz, D5: 1.953125-3.90625Hz is the main range of motion artifacts); then the adaptive thresholding method is used for denoising, and the first filtered signal is reconstructed.

[0101] Kalman adaptive filtering: =0.2 is used as the basis for dynamic adjustment. The process noise matrix is ​​updated, and the second filtered signal is obtained after filtering is completed.

[0102] Dynamic weighted fusion algorithm: Executes the weighted fusion formula, because =0.2, =0.8≥70%, therefore the first filtered signal dominates, and the pulse wave details are preserved first, resulting in the fused filtered signal.

[0103] The second-level subsystem processes the fused and filtered signal, which then enters the second-level subsystem VMD variational mode decomposition: It is decomposed into 6 modal components, of which the 6th mode is a dedicated channel for breathing interference (center angular frequency constrained to 0.1Hz~0.3Hz). The breathing interference modal component was removed, and the remaining 5 modal components were reconstructed to obtain a pure pressure signal stripped of multi-source interference.

[0104] Blood Pressure Measurement and Result Display: Pure pressure signal input blood pressure measurement algorithm: Extract the characteristic envelope of amplitude variation with airbag pressure to determine the maximum peak amplitude. ; Calculate the amplitude corresponding to the systolic pressure according to the preset proportional coefficient. =0.5 Diastolic pressure corresponding amplitude =0.8 Matching to obtain systolic blood pressure =122 mmHg, diastolic blood pressure =78 mmHg; The watch's main control chip sends the blood pressure measurement results to the watch's display screen, which ultimately shows "Systolic pressure: 122 mmHg, Diastolic pressure: 78 mmHg".

[0105] If the user is running (intense exercise level), If the value is 0.6 > 0.5, the measurement will be triggered. The interference intensity classification unit will determine that it is not in the preset interference intensity range, directly trigger a pause measurement prompt, and display "Current motion intensity is too high, please stop and try again" on the watch display to avoid invalid measurements.

[0106] The anti-interference algorithm in this invention achieves a key breakthrough by addressing the core pain points of poor adaptability and inaccurate interference stripping in traditional anti-interference algorithms, with the following advantages: 1. Achieve dynamic adaptive suppression of motion interference. The interference level is quantified by the motion interference intensity coefficient β. The weight allocation and core parameters of the wavelet-Kalman fusion filter are dynamically adjusted to solve the defects of traditional single filtering algorithms or fixed parameter filtering in over-smoothing signals under weak interference and insufficient suppression under strong interference. It significantly improves the signal fidelity and anti-interference robustness under different intensity motion interference scenarios. The signal-to-noise ratio of the signal under motion interference scenarios is significantly improved compared with traditional algorithms.

[0107] 2. Achieve precise and targeted removal of respiratory interference. By locking the range of human respiratory frequency and constraining the center frequency of the dedicated modal channel for respiratory interference in the VMD algorithm, and combining a dynamic penalty factor to avoid modal aliasing, the problem of incomplete removal of respiratory interference and easy aliasing with pulse wave signals by traditional VMD, low-pass filtering and other algorithms is solved. The respiratory interference removal rate is increased to over 97%, and the remaining interference accounts for ≤3%, providing high-purity pulse wave signals for blood pressure calculation.

[0108] 3. It has the ability to collaboratively process complex interference. It adopts a layered anti-interference architecture of "adaptive filtering of motion / vibration interference + directional stripping of respiratory interference" to specifically handle high-frequency sudden interference and low-frequency periodic interference. It solves the shortcomings of traditional single algorithm or serial splicing algorithm in dealing with multi-source complex interference at the same time. The blood pressure measurement error in complex interference scenarios is reduced by more than 35% compared with traditional algorithms, and it is suitable for the measurement needs of complex wrist wearing scenarios.

[0109] 4. The algorithm is lightweight and adapted to smartwatch hardware. All parameter calculation and iteration processes are adapted to the computing power of the smartwatch MCU. The single-round anti-interference processing time is short and the user has no obvious perception. While ensuring the anti-interference effect, it does not affect the measurement real-time performance, solving the problem that traditional complex anti-interference algorithms have high computing power requirements and are difficult to implement on wearable devices. Example

[0110] Please see Figure 4 The present invention provides a signal anti-interference processing system for assisting smartwatches in measuring blood pressure, comprising: The instruction response module 401 is used to respond to the received blood pressure measurement instruction and acquire the initial pulse wave signal of the target user. The data processing module 402 is used to acquire the motion state data of the target user and calculate the motion interference intensity coefficient using the motion state data; The filtering and fusion module 403 is used to perform multi-layer filtering and fusion based on the initial pulse wave signal and the motion interference intensity coefficient when the motion interference intensity coefficient is within a preset interference intensity range, so as to obtain a fused filtered signal. The mode decomposition module 404 is used to perform variational mode decomposition using the fused filtered signal to obtain the target pulse wave signal; The measurement output module 405 is used to convert the target pulse wave signal into blood pressure and obtain the blood pressure measurement result of the target user.

[0111] Since the above is a system corresponding to a signal anti-interference processing method for assisting smartwatches in measuring blood pressure, and its implementation principle is the same as that of a signal anti-interference processing method for assisting smartwatches in measuring blood pressure, for the sake of convenience and brevity, those skilled in the art can clearly understand that the specific working process of the system and modules described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here. Example

[0112] An electronic device according to an embodiment of the present invention includes: a memory and a processor, wherein the memory stores a computer program; when the computer program is executed by the processor, the processor performs a signal anti-interference processing method for assisting a smartwatch in measuring blood pressure as described in any of the above embodiments.

[0113] The memory can be an electronic memory such as flash memory, EEPROM (Electrically Erasable Programmable Read-Only Memory), EPROM, hard disk, or ROM. The memory has storage space for program code used to perform any of the method steps described above. For example, the storage space for program code may include individual program codes for implementing the various steps in the methods described above. This program code can be read from or written to one or more computer program products. These computer program products include program code carriers such as hard disks, compact discs (CDs), memory cards, or floppy disks. The program code may be compressed, for example, in a suitable form. When run by a computing processing device, this code causes the computing processing device to perform the various steps in the methods described above. Example

[0114] This invention provides a computer-readable storage medium storing a computer program thereon, which, when executed, implements the signal anti-interference processing method for assisting smartwatches in measuring blood pressure according to any of the above embodiments.

[0115] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0116] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0117] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0118] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0119] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0120] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A signal anti-interference processing method for assisting smartwatches in measuring blood pressure, characterized in that, include: In response to a blood pressure measurement command, the system acquires the initial pulse wave signal of the target user. Acquire the motion state data of the target user, and use the motion state data to calculate the motion interference intensity coefficient; When the motion interference intensity coefficient is within a preset interference intensity range, multi-layer filtering and fusion are performed based on the initial pulse wave signal and the motion interference intensity coefficient to obtain a fused filtered signal. The target pulse wave signal is obtained by performing variational mode decomposition on the fused filtered signal. The target pulse wave signal is used to convert blood pressure to obtain the blood pressure measurement result of the target user.

2. The signal anti-interference processing method for assisting smartwatches in measuring blood pressure according to claim 1, characterized in that, The step of acquiring the motion state data of the target user and calculating the motion interference intensity coefficient using the motion state data includes: Acquire the motion state data of the target user, wherein the motion state data is the change in three-axis acceleration; Calculate the Euclidean norm of the triaxial acceleration change corresponding to the aforementioned triaxial acceleration change; The motion disturbance intensity coefficient is obtained by calculating the ratio between the Euclidean norm of the triaxial acceleration change and the preset maximum disturbance acceleration.

3. The signal anti-interference processing method for assisting smartwatches in measuring blood pressure according to claim 1, characterized in that, When the motion interference intensity coefficient is within a preset interference intensity range, multi-layer filtering and fusion are performed based on the initial pulse wave signal and the motion interference intensity coefficient to obtain a fused filtered signal, including: When the motion interference intensity coefficient is within the preset interference intensity range, the initial pulse wave signal is used to perform wavelet transform to obtain the first filtered signal. The motion interference intensity coefficient and the initial pulse wave signal are used to perform Kalman adaptive filtering to obtain the second filtered signal; The first filtered signal and the second filtered signal are weighted together with the motion interference intensity coefficient to obtain the fused filtered signal.

4. The signal anti-interference processing method for assisting smartwatches in measuring blood pressure according to claim 3, characterized in that, The step of performing Kalman adaptive filtering using the motion interference intensity coefficient and the initial pulse wave signal to obtain a second filtered signal includes: Determine the Kalman filter parameters, which include the state vector, state transition matrix, observation matrix, observation noise variance, and basic process noise matrix, and initialize the state vector and covariance matrix of the previous time step. The basic process noise matrix is ​​updated using the motion disturbance intensity coefficient to obtain an adaptive process noise matrix; The predicted state vector is obtained by multiplying the state transition matrix with the state vector of the previous time step. The prediction covariance matrix is ​​determined using the state transition matrix, the covariance matrix of the previous time step, the transpose of the state transition matrix, and the adaptive process noise matrix. The Kalman gain is calculated using the prediction covariance matrix, the observation matrix, and the observation noise variance. The state vector at the current moment is obtained by updating the state vector using the Kalman gain, the observation matrix, and the initial pulse wave signal. The first component is extracted from the state vector at the current moment as the second filtered signal.

5. The signal anti-interference processing method for assisting smartwatches in measuring blood pressure according to claim 1, characterized in that, The step of performing variational mode decomposition on the fused filtered signal to obtain the target pulse wave signal includes: Determine the variational mode decomposition parameters; The variational mode decomposition parameters and the fused filtered signal input are used to perform decomposition using a preset variational mode decomposition iterative formula to obtain multiple decomposed mode components; Remove decomposition mode components that fall within a preset breathing frequency range from the multiple decomposition mode components; The remaining multiple decomposed mode components are reconstructed to obtain the target pulse wave signal.

6. The signal anti-interference processing method for assisting smartwatches in measuring blood pressure according to claim 1, characterized in that, The step of converting the target pulse wave signal into blood pressure to obtain the blood pressure measurement result of the target user includes: Extract the characteristic envelope of amplitude variation with wrist airbag pressure from the target pulse wave signal; Based on preset systolic and diastolic blood pressure thresholds, corresponding airbag pressure points are matched from the feature envelope and used as systolic and diastolic blood pressure measurements, respectively. By integrating the systolic blood pressure measurement value and the diastolic blood pressure measurement value, the blood pressure measurement result of the target user is obtained.

7. The signal anti-interference processing method for assisting smartwatches in measuring blood pressure according to any one of claims 1-6, characterized in that, Also includes: When the motion interference intensity coefficient is not within the preset interference intensity range, a blood pressure measurement pause command is generated and fed back to the target user.

8. A signal anti-interference processing system for assisting smartwatches in measuring blood pressure, characterized in that, include: The instruction response module is used to respond to received blood pressure measurement instructions and acquire the initial pulse wave signal of the target user. The data processing module is used to acquire the motion state data of the target user and use the motion state data to calculate the motion interference intensity coefficient. The filtering and fusion module is used to perform multi-layer filtering and fusion based on the initial pulse wave signal and the motion interference intensity coefficient when the motion interference intensity coefficient is within a preset interference intensity range, so as to obtain a fused filtered signal. The mode decomposition module is used to perform variational mode decomposition on the fused filtered signal to obtain the target pulse wave signal; The measurement output module is used to convert the target pulse wave signal into blood pressure and obtain the blood pressure measurement result of the target user.

9. An electronic device, characterized in that, It includes a memory and a processor, wherein the memory stores a computer program capable of being loaded by the processor and executing the signal anti-interference processing method for measuring blood pressure using an auxiliary smartwatch as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer program stores a signal anti-interference processing method for assisting a smartwatch in measuring blood pressure, which can be loaded by a processor and executed as described in any one of claims 1 to 7.

Images