A three-stage measurement state machine home arteriosclerosis detection method, device and storage medium

Abstract

This invention relates to a three-stage measurement state machine method, device, and storage medium for home arteriosclerosis detection, belonging to the field of home health monitoring technology. To address the problem of poor measurement stability in home settings, this method utilizes dual-wavelength PPG and multi-cuff pressure waveforms for collaborative measurement in a home environment, dividing the detection process into three stages: pre-measurement stability assessment, in-measurement dynamic control, and post-measurement recovery analysis. Stage A determines the stability window based on PPG quality and stability; Stage B calculates quality control indicators in real time and performs gating control during cuff inflation and deflation, outputting indicators such as baPWV, ABI, and blood pressure; Stage C continues to collect PPG after depressurization and calculates recovery indicators, generating interpretation labels and retest suggestions. This invention improves the stability and reliability of home arteriosclerosis detection results.

Description

Technical Field

[0001] This invention relates to the field of home health monitoring and non-invasive hemodynamic testing, and in particular to a home arteriosclerosis detection method, device and storage medium based on a three-segment measurement state machine of dual-wavelength PPG and multi-cuff pressure waveform. Background Technology

[0002] Arteriosclerosis-related indicators (such as pulse wave velocity (PWV) and enhancement index (AI)) are of great significance in cardiovascular risk assessment. Current non-invasive measurements in home settings often face problems such as wear differences, motion artifacts, insufficient perfusion, air leakage, and synchronization errors. Common methods only provide a one-time judgment after the measurement, making it difficult for users to correct errors in time during the measurement process. Home-based measurements suffer from insufficient stability and interpretability, and retesting costs are high. Therefore, a closed-loop method is needed that can find a stable window before measurement, dynamically gating during measurement, and providing recovery interpretation and retesting suggestions after measurement. Summary of the Invention

[0003] The purpose of this invention is to provide a home-based method, device, and storage medium for detecting arteriosclerosis using a three-stage measurement state machine. Through stage A (pre-test stability window determination), stage B (real-time quality control gating during test), and stage C (post-test PPG recovery interpretation), a closed-loop process of pre-test stability determination, dynamic control during test, and post-test recovery analysis is achieved, thereby improving the stability and reliability of measurements in home settings.

[0004] To achieve the above objectives, the present invention adopts the following technical solution: In a home setting, the processor controls the PPG acquisition module and the four-channel cuff pressure acquisition module to work collaboratively. In segment A, the quality and stability of the PPG are continuously assessed, and the main cuff measurement is initiated only when the stability window condition is met. In segment B, during cuff inflation / deflation and pressure waveform acquisition, quality control indicators are calculated in real time, and a gating strategy is executed to dynamically adjust the measurement process or output the cause of failure. In segment C, after depressurization, PPG acquisition continues, recovery indicators are calculated, and interpretation labels and retest suggestions are generated and output.

[0005] In one embodiment, the present invention uses a three-stage state machine closed-loop control to divide the measurement process into three stages: pre-measurement stability window determination, real-time quality control gating during measurement, and post-measurement recovery interpretation. During the measurement process, wear differences, motion artifacts, insufficient perfusion, gas leakage, and channel synchronization errors are identified and intervened online, so that measurements that do not meet the conditions are corrected or terminated within the process. At the same time, after the measurement, interpretation tags and retest guidance corresponding to the quality control trigger are generated to achieve the combination of measurement process control and result interpretation. Detailed Implementation

[0006] The embodiments of the present invention will be further described below. It should be understood that the following examples are used to explain the present invention and not to limit the scope of protection of the present invention.

[0007] The home-use arteriosclerosis detection device of the present invention includes: a dual-wavelength PPG acquisition module, a four-channel cuff pressure acquisition module, a pneumatic circuit module, a processor, and a memory. The PPG acquisition module is used to acquire raw PPG signals from both red and infrared wavelengths and output quality-related characteristics. The four-channel cuff pressure acquisition module is used to acquire pressure waveforms from the corresponding cuffs on the left arm, right arm, left ankle, and right ankle, respectively. The pneumatic circuit module includes an inflation pump, a deflation valve, a throttling structure, and a mechanical pressure relief valve. The mechanical pressure relief valve is independently controlled by the processor for safe pressure relief. The processor interacts with a host computer via USB, records event logs, and generates measurement outputs.

[0008] The three-stage measurement state machine includes at least three stages: Stage A, pre-measurement stabilization window; Stage B, measurement gating; and Stage C, post-measurement recovery and interpretation. In Stage A, the device continuously collects PPG data after the user wears the ear clip and calculates the perfusion index (PI), baseline drift, motion artifact index, heart rate stability, and saturation or disconnection status. When these indicators meet thresholds and remain so for a preset time (e.g., 2 to 10 seconds), a stabilization window is established, and the system switches to Stage B. If no window is found within the timeout period, interpretation labels such as "insufficient perfusion," "motion interference," or "poor fit" are output, along with a retest suggestion.

[0009] In segment B, the device initiates the main cuff measurement: after inflation to the target pressure (e.g., 160 mmHg to 200 mmHg), it enters the deflation phase. During deflation, four cuff pressure waveforms are simultaneously acquired, and the oscillation component is extracted from each waveform to calculate quality control indicators such as Q_pressure, synchronization error, leakage characteristics, and consistency of the oscillation envelope. When any quality control indicator triggers a gating condition, the device executes a gating strategy: adjusting the deflation rate or valve duty cycle, adding a supplementary measurement window, reverting to segment A to find a stable window again, or aborting and recording the reason for failure. When the quality control indicator meets the threshold, the device calculates and outputs at least one or more arteriosclerosis-related indicators, such as baPWV, ABI, SBP, DBP, MAP, and home-based AI indicators.

[0010] In one embodiment, Q_pressure is a pressure waveform quality score, ranging from 0 to 100 points. A higher score indicates that the pressure waveform is more suitable for subsequent arrival time and index calculations. Q_pressure is composed of at least platform stability, leakage rate, and oscillation signal-to-noise ratio: the platform is considered stable when the pressure standard deviation σP within the target pressure platform segment is not greater than 1.0 mmHg and the pressure slope |dP / dt| of the platform segment is not greater than 0.5 mmHg / s; simultaneously, the oscillation is considered effective when the signal-to-noise ratio SNR of the oscillation component of the platform segment is not less than 10 dB. In one embodiment, index calculation begins when Q_pressure is greater than or equal to 70 points; a retest or adjustment prompt is triggered when Q_pressure is between 50 and 70 points; and the measurement is terminated and the reason for failure is recorded when Q_pressure is below 50 points.

[0011] In one embodiment, the synchronization error `sync_error` is the maximum deviation at key event moments across multiple channels. Key events include at least the moment of maximum slope of the rising edge of the stroke pressure oscillation or the moment of the envelope peak. In one embodiment, synchronization is considered successful if `sync_error` is no greater than 5ms; if it is greater than 5ms but no greater than 15ms, a synchronization boundary is identified and cross-correlation alignment correction is performed; if it is greater than 15ms, synchronization is considered unsuccessful, and a gating strategy is triggered or a retest is prompted. Cross-correlation alignment correction includes: calculating the time delay Δti corresponding to the maximum cross-correlation value between channels within a preset window, and resampling each channel to a unified time axis to reduce synchronization residuals.

[0012] baPWV can be calculated by combining the arrival time difference Δt of characteristic points of the arm and ankle pressure waveforms with a preset path length model. The path length model can be compensated based on the user's height, age, and body position; it can also compensate for the dependence of blood pressure level, heart rate, and measurement position on PWV or provide prompts in the output.

[0013] In one embodiment, the path length model is used to associate the anatomical path length L of the human body with the user's body surface information. Its input includes at least height H and measurement position; the output is the arm-ankle path length L. In one embodiment, a linear model L = a·H + b is used, where H is in centimeters, L is in millimeters, a is 4.5 to 5.5 mm / cm, and b is 50 to 200 mm. In initial use or verification scenarios, parameters a and b can be individually corrected based on a single control measurement to reduce systematic bias.

[0014] In one embodiment, the home-based AI index is calculated based on the pressure pulse waveform p(t): the first systolic peak P1 and the reflection peak P2 are located within each stroke cycle, and the pulse pressure PP = Psys - Pdia is calculated; the home-based AI is defined as AI = (P2 - P1) / PP × 100%. When the reflection peak cannot be stably located, an alternative reflection intensity index is used, with the ratio of the area of ​​the second half of the systolic segment to the area of ​​the first half or the inflection point amplitude as an approximation of P2, and the alternative mode is marked in the result. In one embodiment, to reduce the influence of heart rate differences, the AI ​​can be normalized to output AI@75 = AI + β·(75 - HR), where β is 0.3 to 0.6% / bpm.

[0015] In segment C, PPG continues to be collected after the cuff is depressurized until the end of the recovery observation period (e.g., 10 to 60 seconds). Recovery time T_rec, perfusion index recovery magnitude ΔPI, recovery slope S_rec, and post-recovery stability are calculated. The device integrates the segment C recovery indicators with segment B quality control trigger information to generate interpretive labels, such as: motion interference, insufficient perfusion, cuff loosening or poor contact, air leakage, synchronization abnormalities, ambient light interference, or arrhythmia risk warnings. It also generates corresponding retesting suggestions, such as retesting while still stationary, readjusting the cuff tightness, changing body position, retesting after warming up, or recommending medical verification.

[0016] In one embodiment, a complete home measurement process includes at least the following steps: S1, the user wears ear clip PPG and completes the wearing of cuffs on all four limbs, and the device detects the connection status and initial pressure of each channel; S2, enters segment A, continuously collects PPG and calculates PI, motion artifact index and heart rate stability, and automatically enters segment B after the stability window reaches a preset duration; S3, enters segment B, the device controls the inflation and deflation of the cuffs and simultaneously collects four pressure waveforms, calculates Q_pressure and sync_error in real time and executes a gating strategy, and outputs baPWV, ABI, SBP, DBP, MAP and home-based AI indicators when the quality control threshold is met; S4, enters segment C, after the cuff is depressurized, PPG is collected again until the end of the observation period, calculates T_rec, ΔPI, S_rec and merges them with the quality control trigger information of segment B to generate interpretation labels and corresponding retest instructions; S5, outputs the measurement results, quality control level and event log identifier through local display or communication interface.

[0017] The present invention can also save the original PPG and pressure data in binary frame format, record event logs and output measurement result files, which facilitates traceability, quality control and remote verification.

Claims

1. A three-stage measurement state machine home arterial stiffness detection method, characterized in that, The method is executed collaboratively by a processor controlling a pressure cuff system and a dual-wavelength photoplethysmography (PPG) sensor in a home setting. The method comprises three segments: segment A (pre-measurement stabilization window), segment B (measurement gating), and segment C (post-measurement recovery and interpretation), with transitions between these segments via a state machine. In segment A, the quality and stability of the PPG signal are assessed, and segment B is only entered when a preset stabilization window condition is met. In segment B, quality control indicators are calculated in real-time during cuff inflation / deflation and pressure waveform acquisition. These indicators include at least a pressure waveform quality score (Q_pressure), synchronization error (sync_error), and leakage characteristics or oscillation envelope consistency. A gating strategy is executed based on the comparison between these indicators and preset thresholds. This gating strategy includes at least dynamic adjustment of the measurement process, additional measurements, or failure assessment. When the quality control indicators meet the threshold conditions, arteriosclerosis-related indicators are calculated and output based on the arrival time difference of characteristic points in the pressure waveform. In segment C, PPG is collected again after cuff depressurization, recovery indicators are calculated, interpretation labels and retest suggestions are generated, and output.

2. The method according to claim 1, characterized in that, The determination of the pre-test stability window of segment A is based on at least one or more of the following: PPG effective perfusion index (PI), pulse amplitude stability, baseline drift, motion artifact index, heart rate stability, or signal saturation / disconnection state; when the indicators meet the threshold and continue to reach the preset time length, the stability window is determined to be established.

3. The method according to claim 1, characterized in that, The B-segment gating includes: controlling the deflation of the cuff at a preset rate or in a closed-loop manner during the deflation phase, and simultaneously acquiring at least four cuff pressure waveforms; separating the pressure waveforms to obtain oscillation components, and calculating the quality score Q_pressure and synchronization error sync_error; triggering the gating strategy when Q_pressure is below the threshold, sync_error exceeds the limit, leakage characteristics appear, or abnormal oscillation envelope appears.

4. The method according to claim 3, characterized in that, The gating strategy includes at least one or more of the following: (1) extending the waiting period in segment A and searching for a stable window again; (2) adjusting the cuff target pressure, inflation rate, deflation rate or deflation valve duty cycle in segment B; (3) adding a supplementary measurement window in segment B or performing supplementary sampling in the failed interval; (4) terminating the current measurement and outputting the reason for failure and retesting suggestions.

5. The method according to claim 1, characterized in that, The arteriosclerosis-related indicators include at least one or more of the following: brachial pulse wave velocity (baPWV), interbrachial pulse wave velocity, ankle-brachial index (ABI), systolic blood pressure (SBP), diastolic blood pressure (DBP), mean arterial pressure (MAP), pressure waveform enhancement index (AI), or their home-based alternative indicators; wherein, baPWV is calculated from the pressure waveforms of the arm and ankle or from the time difference of arrival of characteristic points of the pressure waveform and the PPG waveform combined with a preset path length model.

6. The method according to claim 1, characterized in that, In the post-recovery interpretation of segment C, PPG is collected until the end of the preset recovery observation period, and at least one or more recovery indicators are calculated: recovery time T_rec, perfusion index recovery magnitude ΔPI, recovery slope S_rec, or waveform stability after recovery. Interpretive labels are generated based on recovery indicators and mid-test quality control results. These interpretive labels include at least the following: motion interference, inadequate perfusion, cuff loosening / poor contact, air leakage, synchronization abnormality, ambient light interference, or risk of arrhythmia.

7. The method according to claim 6, characterized in that, The retest recommendations include at least the following: adjusting the wearing position or tightness, remaining still and finding a stable window again, changing the measurement position, retesting after warming or keeping warm, retesting within a specified time period, or recommending medical verification; and the retest recommendations correspond one-to-one with the explanation labels.

8. A home-based arteriosclerosis detection device implementing the method of any one of claims 1 to 7, characterized in that, The device includes a dual-wavelength PPG acquisition module, a four-channel cuff pressure acquisition module, a charge / discharge gas path module, a processor, and a memory. The processor is communicatively connected to both the dual-wavelength PPG acquisition module and the four-channel cuff pressure acquisition module, and is used to receive PPG data and pressure waveform data and execute a three-stage measurement state machine: in stage A, it determines whether to enter stage B based on the PPG stability window; in stage B, it performs gating based on the comparison result of the quality control index and the preset threshold and outputs the index; in stage C, it calculates the PPG recovery index and outputs the interpretation label and retest suggestion. The memory is used to store raw PPG and pressure data, event logs, and measurement output results.

9. The apparatus according to claim 8, characterized in that, The gas circuit module includes an air pump, an air release valve, a throttling structure, and a mechanical pressure relief valve; the mechanical pressure relief valve is used to automatically release pressure when the pressure exceeds a preset safety threshold and is controlled independently of the processor.

10. A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method of any one of claims 1 to 7.