A method, system and device for dynamic assessment of vascular baroreflex sensitivity based on a multi-modal fingertip monitor

Abstract

The application discloses a kind of based on multi-modal fingertip monitor's vascular pressure reflex sensitivity dynamic evaluation method, system and equipment, belong to the technical field of autonomous nerve function assessment of computational physiology.The method relies on monitoring equipment and its specific high-fidelity recording mode that can produce time strictly aligned multi-modal raw data stream in hardware level, relies on synchronous data packet with uniform timestamp that is composed of electrophysiological data sub-packet and optical data sub-packet in data level.The core of method includes: receiving and analyzing the time-aligned data packet, detecting the starting point of skin conductance response event from it, and calculating pulse wave conduction time sequence synchronously;With each skin conductance response event starting point as time reference, analyze the dynamic change characteristics of the pulse wave conduction time sequence in the preset time window after it, and quantify the change characteristics as the evaluation index of vascular pressure reflex sensitivity.The present application makes full use of the accurate timing relationship between skin electrical activity surge event and vascular dynamics instantaneous response provided by high-precision hardware synchronization, solves the problem that traditional pressure reflex sensitivity evaluation method is complex, cannot be continuously monitored, and is difficult to capture instantaneous reflex function.

Description

Technical Field

[0001] This invention relates to the fields of biomedical engineering and digital health, specifically to a method for non-invasively and dynamically assessing cardiovascular autonomic nervous system regulation function—particularly vascular baroreflex sensitivity—using high-precision, multimodal synchronous physiological signals, a system for implementing the method, and specific equipment on which it depends. Background Technology

[0002] Baroreflex sensitivity (BRS) is a key indicator for evaluating the autonomic nervous system's ability to regulate the cardiovascular system, and its impairment is an important marker of the development and progression of diseases such as hypertension and heart failure. Traditional BRS assessment methods, such as invasive pharmacological methods (injection of vasoactive drugs) or sequential methods (analysis of stroke-by-stroke changes in heart rate and blood pressure), have significant limitations: invasive methods are complex and high-risk, making them unsuitable for routine monitoring; non-invasive sequential methods require subjects to remain in a stable state for extended periods, failing to capture transient changes in reflex function during daily activities or under micro-stress, and their results are easily affected by factors such as respiration and movement.

[0003] In recent years, some studies have attempted to use electrodermal activity (EDA) as a "natural stimulus" for sympathetic nerve activation, combined with heart rate response, to assess the skin conductance response (BRS). However, these methods generally face a fundamental technical bottleneck: the data synchronization accuracy of most existing wearable devices is insufficient. There is a delay of hundreds of milliseconds between the surge in EDA (skin conductance response, SCR) and the heart rate response. If there are software synchronization errors of tens of milliseconds or even greater, it will be impossible to accurately establish the instantaneous causal relationship between the two, resulting in low reliability of the calculation results and making it difficult to use for accurate physiological function assessment.

[0004] It must be built on a hardware platform that can provide high-precision time-synchronized physiological data in order to overcome the core defect of insufficient timing accuracy in existing technologies. Summary of the Invention

[0005] To overcome the shortcomings of existing technologies, this invention proposes a method, system, and device for dynamic assessment of vascular baroreflex sensitivity based on a multimodal fingertip monitor. The core of this invention lies in proposing a novel framework for instantaneous BRS assessment that features "patterned high-precision triggering, hardware-level synchronization assurance, and multimodal signal fusion analysis."

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] In a first aspect, the present invention provides a method for assessing vascular baroreflex sensitivity based on multimodal synchronous data, characterized in that the input data of the method is a time-aligned multimodal data packet generated by a synchronous acquisition system operating in its high-fidelity recording mode; the method includes:

[0008] Synchronous data extraction steps: Extract electrophysiological data sub-packets and optical data sub-packets from the multimodal data packets. The electrophysiological data sub-packets contain electrocardiogram signals and skin conductance signals, and the optical data sub-packets contain at least one photoplethysmography (PPG) wave signal with a wavelength sensitive to vascular tension.

[0009] Detection steps: Detect the start time point of at least one skin conductance response event from the skin electrical activity signal; and detect the heartbeat R wave sequence from the electrocardiogram signal;

[0010] Response sequence calculation steps: Based on the heartbeat R-wave sequence and the photoplethysmography pulse wave signal, calculate the pulse wave conduction time corresponding to each heartbeat to form a pulse wave conduction time sequence;

[0011] Dynamic sensitivity quantification steps: For the start time of each skin conductance response event, a preset post-event analysis time window is defined; the change characteristics of the pulse wave conduction time series within the post-event analysis time window relative to the pre-event baseline state are analyzed, and the change characteristics are quantified into a vascular pressure reflex sensitivity index.

[0012] Furthermore, the wavelength of the photoplethysmography signal is selected from a band sensitive to changes in blood oxygen and vascular volume, preferably 850±20nm or 940±20nm.

[0013] Furthermore, the response sequence calculation step specifically involves: for each heartbeat, calculating the time interval between the peak point of the R wave in the electrocardiogram and the characteristic point of the rising branch of the photoplethysmography pulse wave signal waveform, which is taken as the pulse wave conduction time of that heartbeat.

[0014] Furthermore, the dynamic sensitivity quantification step specifically includes: calculating the minimum value of pulse wave conduction time within the post-event analysis time window, and calculating the difference between the minimum value and the average value of pulse wave conduction time within the pre-event baseline time window, as the vascular pressure reflex sensitivity index; and / or, performing trend fitting on the pulse wave conduction time series within the post-event analysis time window, and using the slope obtained from the fitting as the vascular pressure reflex sensitivity index.

[0015] Furthermore, the method also includes a signal quality control step: based on the motion data sub-packets extracted from the multimodal data packet, it is determined whether there is significant motion interference within a preset time period before and after the start time of the skin conductance response event; if so, the calculation results based on the event are discarded.

[0016] Furthermore, the method also includes an environmental compensation step: based on the temperature data sub-packet extracted from the multimodal data packet, a temperature parameter reflecting the peripheral vascular state is obtained, and the parameter is used to compensate and correct the pulse wave conduction time series or the vascular pressure reflection sensitivity index.

[0017] Furthermore, the execution of the method relies on a finger-loop sensor integration structure to provide a stable, high-quality source of electrocardiogram, electrical skin activity, and photoplethysmography (PPG) signals.

[0018] Secondly, the present invention provides a dynamic assessment system for vascular baroreflex sensitivity, characterized in that it comprises:

[0019] The data acquisition interface is used to connect to an acquisition system that can generate time-aligned multimodal data packets and acquire the multimodal data packets generated by it in high-fidelity recording mode.

[0020] The processing module is configured to perform any of the methods described above;

[0021] The output module is used to output the vascular pressure reflex sensitivity index.

[0022] Furthermore, the system is integrated into a remote data analysis platform, and the data acquisition interface reads the multimodal data packets from the storage medium of the acquisition system via wired or wireless means.

[0023] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, characterized in that the program, when executed by a processor, implements the steps of any of the methods described above.

[0024] Beneficial effects:

[0025] 1. Unprecedented Transient BRS Assessment Capability: By using skin conductance response (SCR) events as precise time markers and combining them with rapid vascular response in pulse wave conduction time (PWTT), this invention quantifies transient vascular pressure reflex efficacy during daily micro-stress or autonomic regulation. This overcomes the limitations of traditional methods that can only assess steady-state or long-term average function, providing a new tool for understanding the dynamic characteristics of autonomic nervous system function.

[0026] 2. High Reliability and Accuracy Driven by Data: The effectiveness of this method is strictly based on hardware-level high-precision synchronization of the input data (synchronization error <1 millisecond). This ensures the precise temporal relationship between the "stimulus" (EDA event) and the "response" (PWTT change) signal, making the analysis of instantaneous causal relationships possible and significantly improving the reliability and physiological significance of the evaluation results. Simultaneously, artifact removal and compensation using synchronized motion and temperature data further enhance the robustness of the results.

[0027] 3. Forming a complete technological ecosystem barrier: The advanced analysis method proposed in this invention is deeply coupled with the underlying hardware platform capable of providing specific quality data (i.e., "time-aligned multimodal data packets") and its specific operating mode ("high-fidelity recording mode"). If competitors do not possess hardware systems capable of outputting synchronous data streams with equivalent precision, they cannot effectively implement the method of this invention, thus constructing a complete technological barrier from hardware to algorithms, and from data to insights.

[0028] 4. Directly serving high-value medical and scientific research scenarios: This method can achieve non-invasive and convenient dynamic monitoring, which is very suitable for the efficacy evaluation of antihypertensive drugs or drugs that regulate autonomic nerve function, cardiovascular risk stratification and long-term management of patients with chronic diseases such as heart failure, and mental stress assessment, and has broad prospects for clinical translation and commercial application. Attached Figure Description

[0029] Figure 1 This is a schematic diagram of the instantaneous vascular pressure reflex assessment system and data flow provided in an embodiment of the present invention.

[0030] Figure 2 This is a schematic diagram of the stimulus-response timing principle and index calculation provided in the embodiments of the present invention.

[0031] Figure 3 This is a schematic diagram of the multimodal signal fusion processing and artifact removal process provided in the embodiments of the present invention.

[0032] Figure 4 This is a schematic diagram of the dynamic evaluation application scenario and system output provided in the embodiments of the present invention. Detailed Implementation

[0033] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, so that those skilled in the art can better understand and implement the present invention. However, the embodiments described are not intended to limit the present invention.

[0034] Please see Figure 1This demonstrates the data processing flow and data stream of the instantaneous vascular baroreflex assessment system of the present invention. The core input of the system is a time-aligned multimodal data packet D11 generated by the multimodal monitoring device D10 in its high-fidelity recording mode. This data packet D11 is uploaded to the vascular baroreflex sensitivity assessment engine D12 (BRS assessment engine).

[0035] The core processing flow within the evaluation engine D12 includes six sequential or partially parallel steps:

[0036] Data alignment step D121: Based on the unified timestamp carried in data packet D11, the raw data streams from different sensors (such as electrocardiogram ECG, electrical activity of skin EDA, photoplethysmography pulse wave PPG, motion, and temperature) are precisely time-aligned to ensure that the sampling points of all signals correspond one-to-one on the time axis, which is the basis for subsequent accurate analysis.

[0037] Event detection step D122: Accurately identify skin conductance response (SCR) events from the aligned EDA signal. Each SCR event is characterized by its fast rising edge, and the algorithm (e.g., based on a first derivative threshold) can detect its precise start time point (t_SCR).

[0038] Response calculation step D123: Detect the R-wave peak point of each heartbeat from the synchronized ECG signal, and simultaneously detect characteristic points (such as the foot point of the rising limb of the waveform) from the PPG signal waveform (e.g., 850nm wavelength) sensitive to vascular tension. For each heartbeat, calculate the time interval from the R-wave peak point to the corresponding PPG characteristic point, as the pulse wave conduction time (PWTT) of that heartbeat, thus forming a beat-by-beat PWTT time series.

[0039] Artifact Removal Step D124: Combine the synchronously acquired motion data to perform quality control on each detected SCR event. For example, determine whether the accelerometer signal amplitude exceeds a preset threshold within a certain time window (e.g., ±2 seconds) before and after the event's inception point t_SCR. If it does, the SCR event is considered likely caused by limb movement rather than genuine autonomic nerve activation, and is marked as invalid and removed.

[0040] Step D125 of the index calculation: For each "valid" SCR event that passes quality control, perform core analysis. Define a post-event analysis time window (e.g., t_SCR + 0.5 seconds to t_SCR + 3.0 seconds) with the aforementioned t_SCR as the time zero point. Analyze the variation characteristics of PWTT sequence within this time window relative to the pre-event baseline time window (e.g., t_SCR - 3.0 seconds to t_SCR - 0.5 seconds). For example, calculate the minimum PWTT value PWTT_min within the window and the average PWTT value PWTT_base within the baseline window. The difference between the two, ΔPWTT = PWTT_base - PWTT_min, is quantified as the transient vascular pressure reflex sensitivity index for this event.

[0041] Statistical output step D126: Statistically aggregate the ΔPWTT values ​​calculated from all valid SCR events in a single measurement record, such as calculating the median or average, as a comprehensive indicator representing the "vascular baroreflex response intensity" of this measurement. Finally, this indicator and detailed intermediate results are generated into an evaluation report D13 through the output module and delivered to the clinical application or research analysis end D14.

[0042] Please see Figure 2 This section elaborates on the core physiological principles and computational logic of this method. The figure shows a unified timeline D20, using a single SCR event as an example.

[0043] In the stimulus signal section, a significant rise appears on the EDA signal waveform D21, which is identified as an SCR event D22. The precise starting point D23 of this event is marked as the time zero point t_SCR.

[0044] In the response signal section, the pulse-time series PWTT calculated per beat is shown in D24. A baseline window D25 (pre-event) and a response analysis window D26 (post-event) are defined with t_SCR as the boundary. Physiologically, sympathetic activation (manifested as elevated EDA) causes peripheral vasoconstriction, leading to an increase in arterial pulse wave velocity, thus shortening the PWTT. Therefore, within the response window D26, the PWTT sequence typically shows a decreasing trend, reaching a minimum value PWTT_min D27. Simultaneously, the average value PWTT_base within the baseline window D25 is calculated in D28.

[0045] By calculating ΔPWTT = PWTT_base - PWTT_min D29, a quantified instantaneous response intensity, namely the instantaneous BRS index D210, is obtained. The magnitude of this index directly reflects the sensitivity of the vascular system to the reflexive contractile response to sympathetic nerve stimulation.

[0046] Please see Figure 3 It details the calculation process from raw data segments to final indicators in the form of flowcharts, with particular emphasis on multimodal fusion and strict quality control.

[0047] The process begins with inputting a raw data segment synchronized over a period of time. Step 1, D31, performs initial screening and precise alignment of the raw signal quality. Step 2, D32, performs multimodal feature extraction in parallel: a) detecting SCR events and amplitudes from EDA; b) detecting R waves from ECG; c) calculating pulse-wise time-to-weight ratio (PWTT) from PPG; d) calculating motion amplitude from acceleration signals; e) acquiring temperature data (e.g., fingertip-finger base temperature difference ΔT).

[0048] Step 3, D33, is the crucial quality control and artifact removal stage. It first determines whether the amplitude of the SCR event exceeds the physiologically valid threshold (Condition 1, D331). If it is too weak, it is discarded (Discard 1, D332). For events with sufficient amplitude, it further determines whether there is significant motion interference before and after their occurrence (Condition 2, D333). If so, it is considered a motion artifact and discarded (Discard 2, D334). Only events that pass both conditions are marked as 'valid SCR events', D335.

[0049] Step 4, D34, performs core analysis on valid events. Optionally, the PWTT baseline is first compensated based on temperature data to eliminate the inherent effect of temperature on vascular tone (sub-step a, D341). Then, ΔPWTT is calculated for each valid event (sub-step b, D342). Finally, the ΔPWTT of all valid events in a single measurement is aggregated, and a statistical value (such as the median) is calculated as a comprehensive indicator for this measurement (sub-steps c, D343, and d, D344).

[0050] Step 5, D35, is responsible for outputting and visualizing the results, and finally ends the process.

[0051] Please see Figure 4 This demonstrates the complete workflow and system output of the present invention in a specific application scenario (drug efficacy monitoring).

[0052] Patients or subjects receive alerts or actively trigger measurements via a dedicated application D41 (APP). The multimodal monitoring device D10 (such as a finger cot) is then activated and enters high-fidelity recording mode to collect data. The collected raw data packets are wirelessly uploaded to the cloud analysis platform D42.

[0053] The cloud platform D42 analyzes the data according to the predetermined data collection protocol D43 (e.g., "daily fixed-time measurement", "comparison before and after medication", "continuous monitoring for several weeks"). The analysis results are presented in the form of a long-term trend diagram D44, showing the change curve of the BRS index over time (e.g., number of weeks). For example, it can be clearly observed that the index shows a systematic decline after medication, indicating that the drug is effective.

[0054] On the other hand, a structured and detailed report (D45) is generated for each individual measurement. This report typically includes: a header (patient ID, measurement time, data quality score), key results (such as the number of valid events detected, the average ΔPWTT value and its physiological significance), and visualizations (such as a SCR-PWTT time-series correlation plot). Finally, the report and trend charts are pushed to clinicians or researchers (D46) to provide objective and quantitative decision support for evaluating efficacy and adjusting treatment plans, thus forming a complete closed loop from data collection to clinical intervention.

[0055] Example: Taking the evaluation of the efficacy of a novel antihypertensive drug as an example.

[0056] 1. Data Collection: Recruited subjects sat in a quiet environment every morning before and 2 hours after taking the medication, and activated the "5-minute resting assessment" mode of their finger-cuff monitoring device D10 via the APP. The device automatically entered high-fidelity recording mode, collecting synchronous data such as ECG, EDA, 850nm PPG, acceleration, and temperature.

[0057] 2. Data Processing and Analysis: Data is automatically uploaded to the cloud platform D42. The platform executes as follows: Figure 3 The procedure is illustrated below. For example, in a post-drug measurement, eight valid SCR events (amplitude > 0.05 μs and no motion interference) were detected. For each event, its ΔPWTT was calculated. The median of the eight ΔPWTT values ​​is assumed to be 15.2 milliseconds.

[0058] 3. Indicator Interpretation and Trend Analysis: The platform recorded the "vascular baroreflex intensity" index measured in this study as 15.2 ms. Simultaneously, the baseline measurement index for the subject one week prior to medication (e.g., median of 25.4 ms) was retrieved. The comparison revealed that the index decreased by approximately 40% after medication. This decrease can be explained by the fact that the drug may have reduced the excessive contractile response of blood vessels to mild sympathetic nerve stimulation by regulating autonomic nerve function or acting directly on blood vessels, thus "desensitizing" the sensitivity of the baroreflex. This is often a benign indicator of the effectiveness of certain antihypertensive drugs.

[0059] 4. Clinical Delivery: The above comparative results and long-term trend graphs are integrated into a report and automatically sent to the investigator's clinical data acquisition system. Investigators can obtain early, objective evidence on the effects of the drug on vascular autonomic nerve function within hours, much earlier than the stabilization period of traditional blood pressure measurements, thus accelerating clinical decision-making and the research process.

[0060] The above-described embodiments are merely preferred embodiments to fully illustrate the present invention, and the scope of protection of the present invention is not limited thereto. Equivalent substitutions or modifications made by those skilled in the art based on the present invention are all within the scope of protection of the present invention. The scope of protection of the present invention is defined by the claims.

Claims

1. A method for assessing vascular baroreflex sensitivity based on multimodal synchronous data, characterized in that, The input data for the method is time-aligned multimodal data packets generated by a synchronous acquisition system operating in its high-fidelity recording mode; the method includes: Synchronous data extraction steps: Extract electrophysiological data sub-packets and optical data sub-packets from the multimodal data packets. The electrophysiological data sub-packets contain electrocardiogram signals and skin conductance signals, and the optical data sub-packets contain at least one photoplethysmography (PPG) wave signal with a wavelength sensitive to vascular tension. Event detection steps: Detect the start time point of at least one skin conductance response event from the skin conductance activity signal; and detect the heartbeat R-wave sequence from the electrocardiogram signal; Response sequence calculation steps: Based on the heartbeat R-wave sequence and the photoplethysmography pulse wave signal, calculate the pulse wave conduction time corresponding to each heartbeat to form a pulse wave conduction time sequence; Dynamic sensitivity quantification steps: For the start time of each skin conductance response event, a preset post-event analysis time window is defined; the change characteristics of the pulse wave conduction time series within the post-event analysis time window relative to the pre-event baseline state are analyzed, and the change characteristics are quantified into a vascular pressure reflex sensitivity index.

2. The method according to claim 1, characterized in that, The wavelength of the photoplethysmography signal is selected from the band that is sensitive to changes in blood oxygen and vascular volume, preferably 850±20nm or 940±20nm.

3. The method according to claim 1, characterized in that, The response sequence calculation step is as follows: for each heartbeat, calculate the time interval between the peak point of the R wave in the electrocardiogram and the characteristic point of the rising branch of the photoplethysmography pulse wave signal waveform, which is taken as the pulse wave conduction time of that heartbeat.

4. The method according to claim 1, characterized in that, The dynamic sensitivity quantification step specifically includes: calculating the minimum value of pulse wave conduction time within the post-event analysis time window, and calculating the difference between the minimum value and the average value of pulse wave conduction time within the pre-event baseline time window, as the vascular pressure reflex sensitivity index; and / or, performing trend fitting on the pulse wave conduction time series within the post-event analysis time window, and using the slope obtained from the fitting as the vascular pressure reflex sensitivity index.

5. The method according to claim 1, characterized in that, The method further includes a signal quality control step: based on the motion data sub-packets extracted from the multimodal data packet, it is determined whether there is significant motion interference within a preset time period before and after the start time of the skin conduction response event; If it exists, discard the calculation results based on that event.

6. The method according to claim 1, characterized in that, The method further includes an environmental compensation step: based on the temperature data sub-packet extracted from the multimodal data packet, a temperature parameter reflecting the peripheral vascular state is obtained, and the parameter is used to compensate and correct the pulse wave conduction time series or the vascular pressure reflection sensitivity index.

7. The method according to claim 1, characterized in that, The method relies on a finger-loop sensor integration structure to provide a stable, high-quality source of electrocardiogram, electrical skin activity, and photoplethysmography (PPG) pulse wave signals.

8. A dynamic assessment system for vascular baroreflex sensitivity, characterized in that, include: The data acquisition interface is used to connect to an acquisition system that can generate time-aligned multimodal data packets and acquire the multimodal data packets generated by it in high-fidelity recording mode. The processing module is configured to perform the method as described in any one of claims 1 to 7; The output module is used to output the vascular pressure reflex sensitivity index.

9. The system according to claim 8, characterized in that, The system is integrated into a remote data analysis platform, and the data acquisition interface reads the multimodal data packets from the storage medium of the acquisition system via wired or wireless means.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the method as described in any one of claims 1 to 7.

Images