A method, apparatus, equipment, and medium for self-calibrating acquisition parameters of a flexible ultrasonic patch.

By performing multi-parameter signal processing on the flexible ultrasound patch, time synchronization and self-calibration are achieved, which solves the problem of insufficient accuracy in the auxiliary diagnosis of stroke and improves the reliability and accuracy of diagnosis.

CN123058771APending Publication Date: 2026-11-10THE FIRST MEDICAL CENT CHINESE PLA GENERAL HOSPITAL
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611262158.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-19
Publication Date
2026-11-10

AI Technical Summary

Technical Problem

Existing flexible ultrasound patches for stroke auxiliary diagnosis suffer from problems such as insufficient judgment of single blood flow parameters, lack of time synchronization mechanism for multi-source parameters, large individual differences, and interference from low-quality data, resulting in insufficient diagnostic accuracy.

Method used

By acquiring multi-parameter signals and evaluating contact quality, an individual baseline state is established, time synchronization and self-calibration are achieved, low-quality data are eliminated, multi-parameter time-series features are generated, and auxiliary diagnostic scores are calculated.

Benefits of technology

It improves the accuracy and reliability of flexible ultrasound patches in the auxiliary diagnosis of stroke, overcomes the limitations of single parameter judgment, and eliminates the influence of individual differences and low-quality data.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123058771A_ABST
    Figure CN123058771A_ABST
Patent Text Reader

Abstract

This application discloses a method, apparatus, device, and medium for self-calibrating acquisition parameters of a flexible ultrasound patch. This application belongs to the field of clinical testing data processing technology. The method includes: acquiring acquisition data of a target blood vessel continuously acquired by the flexible ultrasound patch, and simultaneously acquiring blood pressure and heart rate parameters; performing contact quality evaluation and marking low-quality data segments; performing unified time stamping, time window matching, and resampling on multi-source parameters that pass the quality evaluation; establishing individual baseline status; comparing real-time monitoring data with individual baseline status to generate multi-parameter time-series features; performing self-calibration based on the time-series features and data quality evaluation results; and outputting the calibrated parameters and scores. This invention, through data quality control, time synchronization, individualized baseline comparison, and weight adjustment, effectively eliminates interference from poor patch contact and individual differences, improving the accuracy and reliability of flexible ultrasound multi-parameter data in the auxiliary diagnosis of stroke.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This application belongs to the field of clinical test data processing technology, specifically relating to a method, device, equipment, and medium for self-calibrating acquisition parameters of a flexible ultrasound patch. Background Technology

[0002] Stroke is a critical illness with high rates of disability and mortality. Its occurrence and development are closely related to factors such as abnormal cerebral blood supply, changes in blood flow in the carotid or vertebral arteries, abnormal fluctuations in blood pressure, and arrhythmias. Clinically, imaging examinations such as cranial CT and MRI are important bases for the diagnosis of stroke, but they are intermittent examinations, making it difficult to achieve continuous dynamic monitoring of the patient's hemodynamic status. Although traditional desktop ultrasound equipment can detect blood flow velocity, vessel diameter, and blood flow spectrum in the neck vessels, it usually relies on doctors holding the probe to complete short-term examinations, making it impossible to continuously observe the dynamic changes in parameters such as blood flow, blood pressure, and heart rate over a long period of time.

[0003] In recent years, flexible ultrasound patches have been increasingly used for hemodynamic monitoring of superficial blood vessels such as the carotid and vertebral arteries due to their advantages of wearability, lightweight design, and continuous monitoring. However, in actual clinical applications, data acquired by flexible ultrasound patches exhibits significant temporal and multi-parameter characteristics, with inconsistent sampling frequencies, amplitudes of variation, and clinical significance among different parameters. Existing systems typically rely on a single time point or single parameter for judgment, lacking a unified time synchronization mechanism and failing to consider individual differences in baseline blood flow status among patients, easily overlooking the dynamic correlation between blood flow, blood pressure, heart rate, and vascular morphology during stroke. Furthermore, factors such as respiration, changes in body position, and poor patch contact often lead to low-quality data entering the diagnostic model, further reducing the accuracy of auxiliary diagnosis. Therefore, there is an urgent need for a solution capable of quality control, time synchronization, and individualized calibration of multiple parameters acquired by flexible ultrasound patches to improve the accuracy and reliability of stroke auxiliary diagnosis. Summary of the Invention

[0004] This application provides a method, device, equipment, and medium for self-calibrating the acquisition parameters of a flexible ultrasound patch. The purpose is to solve the technical problems existing in the application of flexible ultrasound patches in stroke auxiliary diagnosis, such as the inability of a single blood flow parameter or single time point data to comprehensively reflect stroke-related hemodynamic abnormalities, the lack of a unified time synchronization mechanism among multiple source parameters, the large differences in the baseline blood flow status of different patients leading to insufficient accuracy of fixed threshold judgment, the easy introduction of low-quality data due to interference from respiratory or postural changes and poor patch contact, and the lack of an auxiliary diagnostic result output mechanism based on the dynamic changes of multiple flexible ultrasound parameters.

[0005] In a first aspect, embodiments of this application provide a method for self-calibrating the acquisition parameters of a flexible ultrasonic patch, the method comprising: The system acquires the ultrasound echo signal, blood flow spectrum signal, and cross-sectional signal of the target blood vessel continuously collected by the flexible ultrasound patch, and simultaneously acquires blood pressure and heart rate parameters. Based on the ultrasound echo signal, blood flow spectrum signal, and vascular cross-sectional signal, the contact quality of the acquired data is evaluated, and low-quality data segments are marked. The blood flow parameters, vascular cross-sectional parameters, blood pressure parameters, and heart rate parameters that have passed the quality evaluation are uniformly time-stamped, and data matching and resampling are performed according to a preset time window to achieve time synchronization of multi-source parameters; Under relatively stable conditions, individual baseline status is established based on historical data obtained through quality assessment. The individual baseline status includes baseline blood flow velocity, baseline vascular cross-sectional area, baseline blood pressure, and baseline heart rate. The real-time monitoring data is compared with the corresponding individual baseline status, and the relative change amplitude and dynamic deviation of each parameter are calculated to generate multi-parameter time series features. Based on the multi-parameter time-series characteristics and data quality evaluation results, the real-time monitoring parameters are self-calibrated. The self-calibration includes: reducing the weight of low-quality data segments based on the contact quality evaluation results, and calculating the stroke auxiliary diagnosis score based on the multi-parameter time-series characteristics. Output the calibrated parameters and the stroke auxiliary diagnostic score.

[0006] Furthermore, the contact quality evaluation of the collected data includes: At least one of the following was extracted as a quality evaluation index: ultrasound echo intensity, blood flow spectrum continuity, vascular boundary clarity, vascular center stability, patch contact pressure, patch local deformation, and signal-to-noise ratio. The original signal is graded according to the quality evaluation index, and low-quality data segments are marked.

[0007] Furthermore, the marking of low-quality data segments includes: When any quality evaluation indicator is lower than the corresponding preset standard, the data for the corresponding time period is determined to be a low-quality data segment. Low-quality data segments are excluded from subsequent time synchronization, feature construction, and self-calibration steps, and data quality alerts are generated.

[0008] Furthermore, the process of uniformly time-stamping the blood flow parameters, vascular cross-sectional parameters, blood pressure parameters, and heart rate parameters that have passed the quality evaluation, and performing data matching and resampling according to a preset time window, includes: Give each parameter a unified timestamp and map parameters from different sources to the same timeline; Multi-source parameters are statistically analyzed using a sliding window according to a preset time window, where the time window includes 30 seconds, 1 minute, 5 minutes, 10 minutes, 15 minutes, or 30 minutes. Accordingly, the resampling includes: For parameters with different sampling frequencies, time alignment is performed using resampling, interpolation, moving average, or window statistics methods to ensure that each parameter has the same number of sampling points within the same time window.

[0009] Furthermore, the establishment of an individual's baseline state based on historical data obtained through quality assessment, under relatively stable conditions for the subjects, includes: During the preset initialization period after system startup, multi-source parameters are collected from subjects in a resting or stable state. Data that passes the quality evaluation are selected as historical data, and statistical values ​​of the historical data are calculated as the basic state of an individual. The statistical values ​​include the mean, median, or mode. The baseline blood flow velocity includes baseline values ​​for peak systolic flow velocity, baseline values ​​for end-diastolic flow velocity, or baseline values ​​for mean blood flow velocity; the baseline vascular cross-sectional area includes baseline values ​​for vascular diameter or vascular cross-sectional area; the baseline blood pressure includes baseline values ​​for systolic blood pressure and diastolic blood pressure; and the baseline heart rate includes baseline values ​​for resting heart rate.

[0010] Furthermore, the generation of multi-parameter temporal features includes: Extract at least one of the following parameters: mean, maximum, minimum, rate of change, fluctuation amplitude, coefficient of variation, duration of abnormality, mutation index, left and right blood flow asymmetry index, and degree of deviation relative to the individual's baseline state; The multi-parameter time-series features specifically include: The magnitude of decrease in blood flow velocity, the blood flow velocity mutation index, the magnitude of increase in resistance index, the magnitude of increase in pulsatility index, the rate of change in vascular cross-sectional area, the magnitude of blood pressure fluctuation, and abnormal fluctuations in heart rate.

[0011] Furthermore, the calculation of the stroke auxiliary diagnostic score based on the multi-parameter temporal features includes: Convert the degree of anomaly of each type of parameter into a standardized score; The parameters within each time window are assigned different weights based on the data quality score, with the time window with a high data quality score being assigned a higher weight and the time window with a low data quality score being assigned a lower weight or zero weight. or, A comprehensive abnormality score is calculated based on the combined relationships between abnormal blood flow velocity, abnormal vascular resistance, abnormal blood pressure, and abnormal heart rate. When multiple parameters simultaneously exhibit abnormalities within the same time window, and the duration of the abnormality exceeds a preset threshold, the stroke auxiliary diagnostic score is improved. Output diagnostic results, which include low risk, suspicious abnormality, highly suspicious abnormality, or positive auxiliary diagnostic indications.

[0012] Secondly, embodiments of this application provide a self-calibration device for the acquisition parameters of a flexible ultrasonic patch, the device comprising: The data acquisition module is used to acquire the ultrasound echo signal, blood flow spectrum signal and cross-sectional signal of the target blood vessel continuously acquired by the flexible ultrasound patch, and to acquire blood pressure parameters and heart rate parameters simultaneously. The contact quality evaluation module is used to evaluate the contact quality of the acquired data based on the ultrasound echo signal, blood flow spectrum signal and vascular cross-sectional signal, and to mark low-quality data segments. The time synchronization module is used to uniformly time-mark blood flow parameters, vascular cross-sectional parameters, blood pressure parameters, and heart rate parameters that have passed the quality evaluation, and to perform data matching and resampling according to a preset time window to achieve time synchronization of multi-source parameters. The baseline state establishment module is used to establish an individual baseline state based on historical data obtained through quality assessment when the subject is in a relatively stable state. The individual baseline state includes baseline blood flow velocity, baseline vascular cross-sectional area, baseline blood pressure, and baseline heart rate. The feature construction module is used to compare real-time monitoring data with the corresponding individual baseline status, calculate the relative change magnitude and dynamic deviation of each parameter, and generate multi-parameter time series features; The self-calibration module is used to self-calibrate the real-time monitoring parameters based on the multi-parameter time-series characteristics and data quality evaluation results. The self-calibration includes: reducing the weight of low-quality data segments based on the contact quality evaluation results, and calculating the stroke auxiliary diagnosis score based on the multi-parameter time-series characteristics. The output module is used to output the calibrated parameters and the stroke auxiliary diagnostic score.

[0013] Thirdly, embodiments of this application provide a computer device, including a processor and a memory, wherein the memory stores a computer program, and the processor executes the computer program to implement the method described above.

[0014] Fourthly, embodiments of this application provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method described above.

[0015] The technical solution provided in this application eliminates low-quality data segments through contact quality assessment, achieves accurate time synchronization of multi-source parameters by combining a unified time stamp and a preset time window, eliminates individual differences between different patients by utilizing individual baseline status, generates multi-parameter time-series features based on real-time data and baseline comparison, and completes self-calibration through weight reduction processing and scoring calculation. This effectively solves the limitations of single parameter or time point judgment and significantly improves the accuracy and reliability of continuous monitoring data from flexible ultrasound patches in the auxiliary diagnosis of stroke. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating the self-calibration method for the acquisition parameters of the flexible ultrasonic patch provided in this application embodiment; Figure 2 This is a system overall structure block diagram provided in the embodiments of this application; Figure 3 This is a schematic diagram of the flexible patch ultrasonic prototype and probe provided in the embodiments of this application; Figure 4 This is a schematic diagram of flexible ultrasound patch wearing and multi-vessel monitoring provided in the embodiments of this application; Figure 5 This is a schematic diagram of multi-source parameter acquisition and time synchronization provided in an embodiment of this application; Figure 6 This is a schematic diagram of the structure of the self-calibration device for the acquisition parameters of the flexible ultrasonic patch provided in the embodiments of this application. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this application clearer, specific embodiments of this application will be described in further detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely for explaining this application and not for limiting it. It should also be noted that, for ease of description, only the parts relevant to this application are shown in the drawings, not all of them. Before discussing exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although the flowcharts describe operations (or steps) as sequential processes, many of these operations can be performed in parallel, concurrently, or simultaneously. Furthermore, the order of the operations can be rearranged. The process can be terminated when its operation is completed, but may also have additional steps not included in the drawings. The process can correspond to a method, function, procedure, subroutine, subroutine, etc.

[0018] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0019] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such terms can be used interchangeably where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class and the number of objects is not limited; for example, a first object can be one or more. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0020] The following description, in conjunction with the accompanying drawings, details the self-calibration method, apparatus, equipment, and medium for the acquisition parameters of the flexible ultrasonic patch provided in this application, through specific embodiments and application scenarios.

[0021] Example 1 Figure 1 This is a flowchart illustrating the self-calibration method for the acquisition parameters of the flexible ultrasonic patch provided in Embodiment 1 of this application. Figure 1 As shown, the specific steps include the following: S1: Acquire the ultrasound echo signal, blood flow spectrum signal and cross-sectional signal of the target blood vessel continuously acquired by the flexible ultrasound patch, and simultaneously acquire blood pressure parameters and heart rate parameters. A flexible ultrasound patch can be a wearable ultrasound probe that can be attached to the human body surface to non-invasively and continuously acquire vascular ultrasound signals. For example, a flexible ultrasound patch can be attached to the skin of the neck to acquire carotid artery signals.

[0022] The target blood vessel can be any human artery suitable for superficial ultrasound monitoring, specifically the common carotid artery, internal carotid artery, vertebral artery, brachial artery, or radial artery.

[0023] Ultrasound echo signals can be tissue reflection echoes received after ultrasound waves are emitted by a flexible ultrasound patch, and can be used to identify blood vessel boundaries and subcutaneous tissue layers.

[0024] Blood flow spectrum signals can be obtained based on the principle of Doppler ultrasound and represent the spectral signals that characterize the state of blood flow motion inside blood vessels.

[0025] Cross-sectional signals of blood vessels can be ultrasound imaging signals that reflect the cross-sectional shape and diameter of the blood vessel.

[0026] Blood pressure parameters can be physiological parameters that characterize the pressure level of blood vessels in the human body. For example, blood pressure parameters include systolic blood pressure, diastolic blood pressure, and mean arterial pressure.

[0027] Heart rate parameters can be physiological parameters that characterize the frequency of heartbeats and changes in heart rate.

[0028] This solution can simultaneously acquire ultrasound signals and blood pressure and heart rate data streams from external monitoring devices, while reading the continuous raw ultrasound signals output by the flexible ultrasound patch.

[0029] S2: Based on the ultrasound echo signal, blood flow spectrum signal, and vascular cross-sectional signal, the contact quality of the acquired data is evaluated, and low-quality data segments are marked. Contact quality evaluation is a data processing procedure that determines the adhesion status of flexible ultrasonic patches to the skin based on the characteristics of the original ultrasound signals.

[0030] Low-quality data segments can be continuous time series data that are distorted and lack effective information due to factors such as patch displacement, motion interference, and changes in body position.

[0031] Specifically, the obtained ultrasound echo signals, blood flow spectrum signals, and vascular cross-sectional signals can be analyzed to determine the quality of contact and mark the continuous time intervals that are deemed to be of substandard quality.

[0032] S3: The blood flow parameters, vascular cross-sectional parameters, blood pressure parameters, and heart rate parameters that have passed the quality evaluation are uniformly time-stamped, and data matching and resampling are performed according to the preset time window to achieve time synchronization of multi-source parameters; Blood flow parameters can be hemodynamic indices extracted from blood flow spectrum signals. For example, blood flow parameters include peak systolic velocity, end-diastolic velocity, and drag index.

[0033] Blood vessel cross-sectional parameters can be obtained from the analysis of blood vessel cross-sectional signals and are indicators that characterize the morphological features of blood vessels. For example, blood vessel cross-sectional parameters include blood vessel diameter and blood vessel cross-sectional area.

[0034] A preset time window can be a pre-defined continuous duration used for segmented statistical processing of time series data. For example, a preset time window can be 5 minutes.

[0035] Resampling can be a process that transforms the original sampling points, converting data from different sampling frequencies into data with the same sampling frequency.

[0036] Multi-source parameter time synchronization can align multiple physiological parameters from ultrasound patch, blood pressure acquisition module, and heart rate acquisition module to the same time axis, ensuring that multiple parameters can be compared and analyzed within the same time window.

[0037] This solution can add a time stamp to each parameter data point for qualified data that was not marked in the above steps, complete the data matching of parameters from different sources according to a preset time window, and eliminate the sampling frequency difference between devices through resampling processing to achieve time alignment of multi-source parameters.

[0038] S4: Under relatively stable conditions, establish individual baseline status based on historical data obtained through quality assessment. The individual baseline status includes baseline blood flow velocity, baseline vascular cross-sectional area, baseline blood pressure, and baseline heart rate. A relatively stable state for the subject can be defined as a physiological state in which the subject is at rest, without significant physical activity, and with stable emotions. Specifically, this can be a lying or sitting position.

[0039] Historical data can be past monitoring datasets collected in the early stages of the system and verified through contact quality evaluation.

[0040] Individual baseline status can be a personalized physiological benchmark established for the current subject, used for comparison with subsequent real-time monitoring data, including baseline blood flow velocity, baseline vascular cross-sectional area, baseline blood pressure, and baseline heart rate.

[0041] Historical data of subjects in a relatively stable state and that have passed quality verification can be selected and statistically calculated to generate the individual baseline state specific to the subject, which can then be used as a personalized reference benchmark for subsequent judgment of whether parameters are abnormal.

[0042] S5: Compare real-time monitoring data with the corresponding individual baseline status, calculate the relative change amplitude and dynamic deviation of each parameter, and generate multi-parameter time series features; Real-time monitoring data can be the latest physiological monitoring dataset continuously collected by the system.

[0043] The relative change range can be the proportion of the change in the real-time monitoring parameter relative to the individual baseline value.

[0044] The degree of dynamic deviation can be a quantitative representation of how much the real-time monitoring parameters deviate from the individual's baseline state.

[0045] Multi-parameter time-series features can be a set of features extracted from time-series physiological data that can characterize the dynamic changes of parameters over time.

[0046] After reading the real-time monitoring data, it is compared with the individual basic state obtained in the previous steps. The relative change range and dynamic deviation of each parameter are calculated respectively, and multi-parameter time series features are extracted and generated based on the calculation results.

[0047] S6: Based on the multi-parameter time-series characteristics and data quality evaluation results, the real-time monitoring parameters are self-calibrated. The self-calibration includes: reducing the weight of low-quality data segments based on the contact quality evaluation results, and calculating the stroke auxiliary diagnosis score based on the multi-parameter time-series characteristics. Self-calibration is a data processing procedure that combines signal quality and individual physiological benchmarks to perform weight adjustments and risk quantification scoring on real-time monitoring parameters.

[0048] Weight reduction processing can decrease the weight ratio of parameters corresponding to low-quality data segments in subsequent calculations, thereby reducing the interference of distorted data on the results.

[0049] The stroke auxiliary diagnostic score can be a quantitative risk score output by the system. It is only used to provide auxiliary indication of hemodynamic risk of stroke and is not equivalent to a medical diagnosis.

[0050] Specifically, the multi-parameter time-series features and data quality evaluation results can be used to perform parameter-level self-calibration by reducing the weight of low-quality data segments, and to calculate the stroke auxiliary diagnosis score using the multi-parameter time-series features.

[0051] S7: Output the calibrated parameters and the stroke auxiliary diagnostic score.

[0052] The calibrated parameters can be physiological parameter datasets after weight reduction correction and time alignment.

[0053] The self-calibrated parameters and stroke auxiliary diagnostic scores can be output to a display interface or to downstream storage or diagnostic modules.

[0054] This technical solution identifies low-quality data segments through contact quality assessment, performs personalized comparisons based on individual baseline conditions, and introduces a weighted self-calibration method to avoid directly discarding data and causing loss of effective information. It solves the problems of inconsistent sampling from different devices, differences in individual physiological baselines, and interference with monitoring results due to poor patch contact, providing a calibrated time-series parameter basis for the auxiliary assessment of stroke risk.

[0055] In one embodiment, optionally, the contact quality assessment of the collected data includes: At least one of the following was extracted as a quality evaluation index: ultrasound echo intensity, blood flow spectrum continuity, vascular boundary clarity, vascular center stability, patch contact pressure, patch local deformation, and signal-to-noise ratio. The original signal is graded according to the quality evaluation index, and low-quality data segments are marked.

[0056] Among them, the ultrasonic echo intensity can be the amplitude of the tissue echo signal received by the flexible ultrasonic patch, which can reflect the strength of the ultrasonic signal reception.

[0057] Blood flow spectrum continuity can be a measure of whether the blood flow spectrum signal is continuous and uninterrupted over time.

[0058] The clarity of blood vessel boundaries refers to the degree to which the contours of the blood vessel walls are identifiable in ultrasound images.

[0059] Vascular center stability refers to the degree to which the position of the vascular center does not drift within consecutive frames of ultrasound images.

[0060] The patch contact pressure can be the pressure value between the flexible ultrasonic patch and human skin.

[0061] Local deformation of the patch can refer to the degree of local compression, bending, or deformation that occurs after the flexible ultrasonic patch is attached to the body surface.

[0062] Signal-to-noise ratio (SNR) is the ratio of effective signal power to noise power, used to measure the degree of noise interference to a signal.

[0063] Quality evaluation indicators can be a type of quantitative or semi-quantitative indicator used to judge the quality of ultrasound acquisition data.

[0064] Quality grading can be a process that divides the raw signal into different quality levels according to its quality.

[0065] Specifically, at least one piece of information, such as ultrasound echo intensity and blood flow spectrum continuity, can be extracted from the original ultrasound signal and patch sensing information as a quality evaluation index. Then, based on the selected index, the quality level of the original signal is classified, and low-quality data segments are identified and marked.

[0066] This technical solution provides a multi-dimensional set of quality evaluation indicators, no longer relying on a single dimension to judge the quality of data. It can carry out quality control by comprehensively considering signal image features and the deformation and pressure state of the patch itself, thereby improving the comprehensiveness of contact quality evaluation and reducing the situation of misjudging low-quality data segments by a single indicator.

[0067] In one embodiment, optionally, marking low-quality data segments includes: When any quality evaluation indicator is lower than the corresponding preset standard, the data for the corresponding time period is determined to be a low-quality data segment. Low-quality data segments are excluded from subsequent time synchronization, feature construction, and self-calibration steps, and data quality alerts are generated.

[0068] The preset standard can be a pre-set threshold for each quality evaluation indicator, such as the minimum acceptable threshold for signal-to-noise ratio.

[0069] Data quality alerts can be system-generated messages that inform users of monitoring segments that have substandard quality.

[0070] Each quality evaluation indicator is compared with the preset standard corresponding to that indicator; if any indicator fails to meet the preset standard, the time period is judged as a low-quality data segment, and the data in that segment is excluded from the subsequent time synchronization, feature construction, and self-calibration calculation process. At the same time, a corresponding data quality prompt is generated and output.

[0071] This technical solution clarifies the triggering and judgment logic for low-quality data segments and directly excludes severely unqualified data from subsequent calculations, avoiding severely distorted data from interfering with the overall results. At the same time, it outputs prompts to help users understand the source of data anomalies, facilitating the investigation of issues related to patch wearing and body positioning.

[0072] In one embodiment, optionally, the step of uniformly time-stamping the blood flow parameters, vascular cross-sectional parameters, blood pressure parameters, and heart rate parameters that have passed the quality evaluation, and performing data matching and resampling according to a preset time window, includes: Give each parameter a unified timestamp and map parameters from different sources to the same timeline; Multi-source parameters are statistically analyzed using a sliding window according to a preset time window, where the time window includes 30 seconds, 1 minute, 5 minutes, 10 minutes, 15 minutes, or 30 minutes. Accordingly, the resampling includes: For parameters with different sampling frequencies, time alignment is performed using resampling, interpolation, moving average, or window statistics methods to ensure that each parameter has the same number of sampling points within the same time window.

[0073] A timestamp can be a time information that marks the moment when data was collected.

[0074] The time axis can be a reference coordinate system with time as the horizontal axis, used to align various timing parameters.

[0075] Sliding window statistics is a data processing method that involves scrolling the window sequentially over time and performing statistical calculations on the data within the window segment by segment.

[0076] Interpolation can be a data completion algorithm that uses existing discrete sampling points to estimate intermediate time values, such as linear interpolation.

[0077] Moving average is a data smoothing method that calculates the average of a range of sample points.

[0078] Specifically, a unified timestamp can be added to each parameter, and parameters from different acquisition devices can be mapped to the same time axis. Selectable durations such as 30 seconds or 1 minute as preset time windows to carry out sliding window statistics. For parameters with inconsistent sampling frequencies, one or more combinations of resampling, interpolation, moving average, and window statistics can be used to complete time alignment, ensuring that the number of sampling points for all parameters is consistent within the same time window.

[0079] This technical solution refines the complete implementation methods of time synchronization, provides selectable time window ranges, and offers multiple alignment algorithms to address the problem of inconsistent sampling frequencies. It solves the problem of asynchronous sampling rates of ultrasound, blood pressure, and heart rate devices, ensuring that multi-source parameters can correspond one-to-one in the time dimension, and providing a well-organized time-series dataset for subsequent time-series feature calculations.

[0080] In one embodiment, optionally, establishing an individual baseline state based on historical data from quality assessments while the subject is in a relatively stable state includes: During the preset initialization period after system startup, multi-source parameters are collected from subjects in a resting or stable state. Data that passes the quality evaluation are selected as historical data, and statistical values ​​of the historical data are calculated as the basic state of an individual. The statistical values ​​include the mean, median, or mode. The baseline blood flow velocity includes baseline values ​​for peak systolic flow velocity, baseline values ​​for end-diastolic flow velocity, or baseline values ​​for mean blood flow velocity; the baseline vascular cross-sectional area includes baseline values ​​for vascular diameter or vascular cross-sectional area; the baseline blood pressure includes baseline values ​​for systolic blood pressure and diastolic blood pressure; and the baseline heart rate includes baseline values ​​for resting heart rate.

[0081] The preset initialization period can be a fixed duration after the system starts up, specifically used for collecting baseline data. For example, the preset initialization period can be set to 5 minutes.

[0082] Resting can be a physiological state in which the subject's body remains quiet and undisturbed by movement.

[0083] A statistical value can be a representative value obtained by performing statistical operations on a batch of historical data, and can be the mean, median, or mode.

[0084] The baseline value of peak systolic velocity, etc., are all personalized baseline values ​​collected from subjects at rest.

[0085] During the preset initialization period after system startup, multi-source parameters of subjects in a stable resting state can be collected, and data that has passed quality evaluation can be selected as historical data. The mean, median or mode of the historical data can be calculated as the individual's baseline state, and the baseline values ​​of blood flow, blood vessels, blood pressure and heart rate can be obtained respectively.

[0086] This technical solution limits the timing of individual baseline collection, data screening rules, and statistical calculation methods, and refines the specific benchmark parameters contained within the individual's basic state. It generates personalized baselines based on the subject's own initialization phase data, overcoming the shortcomings of large differences in physiological baselines among different subjects and the inaccuracy caused by using a globally fixed threshold.

[0087] In one embodiment, optionally, generating multi-parameter time-series features includes: Extract at least one of the following parameters: mean, maximum, minimum, rate of change, fluctuation amplitude, coefficient of variation, duration of abnormality, mutation index, left and right blood flow asymmetry index, and degree of deviation relative to the individual's baseline state; The multi-parameter time-series features specifically include: The magnitude of decrease in blood flow velocity, the blood flow velocity mutation index, the magnitude of increase in resistance index, the magnitude of increase in pulsatility index, the rate of change in vascular cross-sectional area, the magnitude of blood pressure fluctuation, and abnormal fluctuations in heart rate.

[0088] Rate of change can be a measure of how quickly a parameter changes value per unit of time.

[0089] The fluctuation range can be the range of fluctuation obtained by subtracting the minimum value from the maximum value of the parameter over a period of time.

[0090] The coefficient of variation (CV) is the ratio of the standard deviation to the mean, and it characterizes the level of dispersion and fluctuation in data.

[0091] The duration of an anomaly can be the continuous length of time that a parameter remains within the anomaly range.

[0092] The mutation index can be used to quantify the degree of sudden change in a parameter.

[0093] The left-right blood flow asymmetry index can be used as an indicator to quantify the difference in blood flow parameters between the left and right sides of the human body.

[0094] The magnitude of the decrease in blood flow velocity can be the order of magnitude of the decrease in blood flow velocity compared to the baseline.

[0095] The magnitude of the increase in the resistance index can be the order of magnitude of the increase in the resistance index relative to the baseline.

[0096] The magnitude of the increase in the pulsatility index can be measured by the order of magnitude of the increase in the pulsatility index relative to the baseline.

[0097] The rate of change of blood vessel cross-sectional area can be the proportion of change of blood vessel cross-sectional area relative to the baseline.

[0098] Blood pressure fluctuation range can be defined as the range of blood pressure fluctuations within a time window.

[0099] Abnormal heart rate fluctuations can be a phenomenon in which the heart rate undergoes abnormal and large-scale oscillations within a short period of time.

[0100] At least one feature, such as the mean, maximum, coefficient of variation, mutation index, and left-right blood flow asymmetry index, can be extracted from the time-synchronized time-series parameters. Further analysis can then yield specialized multi-parameter time-series features for stroke monitoring, such as the magnitude of blood flow velocity decrease and the magnitude of resistance index increase.

[0101] This technical solution provides general time-series statistical features as well as stroke-specific feature sets. It includes general time-series statistical indicators and closely matches the clinical manifestations of abnormal cerebrovascular hemodynamics, enriching the feature dimensions. It can capture various risk signals such as parameter mutations, persistent abnormalities, and left-right blood flow imbalances, providing sufficient feature input for subsequent self-calibration and auxiliary diagnostic scoring.

[0102] In one embodiment, optionally, calculating the stroke auxiliary diagnostic score based on the multi-parameter temporal features includes: Convert the degree of anomaly of each type of parameter into a standardized score; The parameters within each time window are assigned different weights based on the data quality score, with the time window with a high data quality score being assigned a higher weight and the time window with a low data quality score being assigned a lower weight or zero weight. or, A comprehensive abnormality score is calculated based on the combined relationships between abnormal blood flow velocity, abnormal vascular resistance, abnormal blood pressure, and abnormal heart rate. When multiple parameters simultaneously exhibit abnormalities within the same time window, and the duration of the abnormality exceeds a preset threshold, the stroke auxiliary diagnostic score is improved. Output diagnostic results, which include low risk, suspicious abnormality, highly suspicious abnormality, or positive auxiliary diagnostic indications.

[0103] Standardized scoring can be a scoring system that maps the degree of abnormality of parameters with different dimensions to a unified score range, such as a score range of 0-100.

[0104] Data quality score can be a quantitative score given for the overall quality of the data within a window.

[0105] The comprehensive anomaly score can be a quantitative score of overall anomalies calculated by combining the relationships between multiple types of anomalies.

[0106] The preset threshold can be a pre-defined minimum duration of an anomaly, for example, the preset threshold can be set to 30 seconds.

[0107] The diagnostic results can be risk level prompts output by the system, which are divided into low risk, suspicious abnormality, highly suspicious abnormality, or positive auxiliary diagnosis prompts.

[0108] Two calculation paths can be selected. The first path converts the degree of abnormality of each parameter into a standardized score and assigns weights based on the data quality score of each window, with higher quality data receiving higher weights. The second path calculates a comprehensive abnormality score based on the abnormal combination of blood flow velocity, vascular resistance, blood pressure, and heart rate. When multiple parameters are abnormal simultaneously within the same time window, and the duration of the abnormality exceeds a preset threshold, the stroke auxiliary diagnostic score is increased, and the corresponding diagnostic result is finally output.

[0109] This technical solution refines the two sets of calculation logic for stroke auxiliary diagnostic scoring, realizes quality-driven weight allocation, and introduces rules for multiple parameters with concurrent abnormalities and continuous score increases exceeding the threshold. It takes into account both data quality and the clinical significance of multi-parameter coupling, and outputs graded risk warnings, making the scoring results more consistent with the occurrence pattern of stroke hemodynamic abnormalities, and the results are interpretable and convenient for clinical personnel to refer to.

[0110] To enable those skilled in the art to better understand this solution, this application also provides a preferred embodiment.

[0111] A stroke diagnostic system based on flexible ultrasound multi-parameter temporal fusion. Figure 2 This is a system overall structure block diagram provided in the embodiments of this application, with reference to... Figure 2 The system includes the following modules: 1. Flexible ultrasonic patch data acquisition module.

[0112] The flexible ultrasound patch data acquisition module is used to acquire ultrasound echo signals, blood flow spectrum signals, and cross-sectional signals of the target blood vessel. The target blood vessel includes at least one of the common carotid artery, internal carotid artery, and radial artery.

[0113] 2. Access the quality evaluation module.

[0114] The contact quality evaluation module is used to evaluate the quality of data acquired by the flexible ultrasound patch based on ultrasound echo intensity, blood flow spectrum continuity, vascular boundary clarity, vascular center stability, patch contact pressure, patch local deformation, and signal-to-noise ratio, and to mark low-quality data segments.

[0115] 3. Blood flow parameter extraction module.

[0116] The blood flow parameter extraction module is used to extract parameters such as peak systolic velocity, end-diastolic velocity, resistance index, pulsatility index, mean blood flow velocity, and blood flow direction from the blood flow spectrum signal.

[0117] 4. Blood vessel cross-sectional parameter extraction module.

[0118] The blood vessel cross-sectional parameter extraction module is used to extract parameters such as blood vessel diameter, blood vessel cross-sectional area, and blood vessel area change rate from blood vessel cross-sectional signals.

[0119] 5. Blood pressure and heart rate parameter acquisition module.

[0120] The blood pressure and heart rate parameter acquisition module is used to collect or receive parameters such as systolic blood pressure, diastolic blood pressure, and heart rate.

[0121] 6. Time synchronization module.

[0122] The time synchronization module is used to uniformly time-mark blood flow parameters, vascular cross-sectional parameters, blood pressure parameters, and heart rate parameters acquired by the flexible ultrasound patch, and to perform data matching and resampling according to a preset time window.

[0123] 7. Multi-parameter temporal feature construction module.

[0124] The multi-parameter temporal feature construction module is used to construct dynamic features within different time windows, including parameter mean, maximum, minimum, rate of change, fluctuation amplitude, coefficient of variation, abnormal duration, mutation index, left and right blood flow asymmetry index, and deviation from the individual's baseline state.

[0125] 8. Time-series fusion diagnostic module.

[0126] The temporal fusion diagnostic module is used to fuse and analyze the temporal characteristics of blood flow parameters, vascular cross-sectional parameters, blood pressure parameters, and heart rate parameters to generate a stroke auxiliary diagnostic score.

[0127] 9. Diagnostic result output module.

[0128] The diagnostic results output module is used to output stroke auxiliary diagnostic results, abnormal hemodynamic indications, sources of abnormal parameters, data quality indications, and suggestions for re-examination.

[0129] The stroke diagnosis system based on flexible ultrasound multi-parameter temporal fusion includes the following steps: Step 1: The subject wears a flexible ultrasound patch.

[0130] The flexible ultrasound patch is attached to the corresponding area of ​​the target blood vessel on the body surface. The system is then activated to collect the ultrasound echo signal, blood flow spectrum signal, and cross-sectional signal of the target blood vessel. Figure 3 This is a schematic diagram of the flexible patch ultrasonic prototype and probe provided in the embodiments of this application. Figure 4 This is a schematic diagram of flexible ultrasound patch wearing and multi-vessel monitoring provided in the embodiments of this application. This solution can perform corresponding data monitoring after the subject has finished wearing the patch.

[0131] Step 2: Collect multiple physiological parameters.

[0132] The system continuously collects blood flow parameters, vascular cross-sectional parameters, blood pressure parameters, and heart rate parameters. Blood flow parameters include peak systolic velocity, end-diastolic velocity, resistance index, pulsatility index, mean blood flow velocity, and blood flow direction; vascular cross-sectional parameters include vessel diameter and cross-sectional area; blood pressure parameters include systolic and diastolic blood pressure; and heart rate parameters include heart rate.

[0133] Step 3: Data quality control.

[0134] The system evaluates the quality of acquired data based on factors such as blood flow spectrum continuity, vessel boundary clarity, vessel center drift, patch contact status, and signal-to-noise ratio. If a data point exhibits low contact quality, interrupted blood flow spectrum, unidentifiable vessel boundaries, or vessel center drift exceeding a preset range within a given timeframe, that timeframe is marked as low-quality data and excluded from diagnostic fusion analysis.

[0135] Step 4: Time synchronization and data matching.

[0136] Figure 5 This is a schematic diagram of multi-source parameter acquisition and time synchronization provided in an embodiment of this application. The system uniformly timestamps parameters from different sources and performs synchronization matching according to a preset time window. The time window can be 30 seconds, 1 minute, 5 minutes, 10 minutes, 15 minutes, or 30 minutes. For parameters with different sampling frequencies, the system can use resampling, interpolation, moving average, or window statistics methods for time alignment.

[0137] Step 5: Establishing the individual's basic state.

[0138] The system establishes an individual's baseline status under relatively stable conditions, including baseline blood flow velocity, baseline vascular cross-sectional area, baseline blood pressure, and baseline heart rate. Subsequent monitoring data are compared with the individual's baseline status to calculate the relative magnitude of change and the degree of dynamic deviation.

[0139] Step 6: Construction of multi-parameter temporal features.

[0140] The system extracts temporal features of blood flow, blood pressure, heart rate, and vascular cross-sectional parameters within different time windows. These temporal features include the magnitude of decrease in blood flow velocity, the blood flow velocity mutation index, the magnitude of increase in the resistance index, the magnitude of increase in the pulsatility index, the rate of change in vascular cross-sectional area, the magnitude of blood pressure fluctuations, and abnormal fluctuations in heart rate.

[0141] Step 7: Multi-parameter temporal fusion diagnosis.

[0142] The system inputs temporal features from multiple time windows into the temporal fusion diagnostic module to calculate a stroke auxiliary diagnostic score. The temporal fusion diagnostic module can employ a rule-based scoring model, a logistic regression model, a random forest model, a gradient boosting tree model, a recurrent neural network model, an attention mechanism model, or a combination of the above models.

[0143] Step 8: Output the diagnostic results.

[0144] The system outputs diagnostic results based on the stroke auxiliary diagnostic score. Diagnostic results may include low risk, suspicious abnormality, highly suspicious abnormality, or a positive indication for stroke auxiliary diagnosis. Simultaneously, the system outputs the sources of key abnormal parameters, such as abnormal blood flow velocity, abnormal vascular resistance, abnormal blood flow asymmetry, abnormal blood pressure fluctuations, abnormal heart rate, or abnormal heart rhythm.

[0145] Step 9: Report generation.

[0146] The system automatically generates a diagnostic report, which includes monitoring time, target blood vessel location, proportion of valid data, proportion of low-quality data, main abnormal parameters, time-series change curve, stroke auxiliary diagnostic score, diagnostic results, and follow-up recommendations.

[0147] The multi-parameter time series fusion metrics described in this invention include at least two of the following categories: 1. Blood flow velocity parameters: peak systolic velocity, end-diastolic velocity, mean blood flow velocity, blood flow velocity mutation index, blood flow velocity decrease rate, and blood flow velocity coefficient of variation; 2. Vascular resistance indicators: resistance index, pulsatility index, rate of change of resistance index, and rate of change of pulsatility index; 3. Vascular morphology indicators: changes in vessel diameter, cross-sectional area, and vessel area; 4. Blood pressure indicators: systolic blood pressure, diastolic blood pressure, mean arterial pressure, and blood pressure fluctuation range; 5. Heart rate indicators.

[0148] The stroke auxiliary diagnostic score described in this invention can be obtained in the following ways: 1. Convert the degree of abnormality of each type of parameter into a standardized score; 2. Assign different weights to parameters within each time window based on the data quality score; 3. Calculate a comprehensive abnormality score based on the combined relationships between abnormal blood flow velocity, abnormal vascular resistance, abnormal blood pressure, and abnormal heart rate; 4. When multiple parameters simultaneously exhibit abnormalities within the same time window, and the duration of the abnormality exceeds a preset threshold, improve the stroke auxiliary diagnostic score; 5. When abnormal data mainly originates from low-quality data segments, reduce the diagnostic weight of the corresponding time window; 6. Output stroke auxiliary diagnosis results and sources of abnormal parameters based on the comprehensive score.

[0149] The stroke auxiliary diagnostic score can be expressed as low risk, suspicious abnormality, highly suspicious abnormality, or positive auxiliary diagnostic indication.

[0150] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention enables unified time synchronization and fusion analysis of blood flow parameters, vascular cross-sectional parameters, blood pressure parameters, and heart rate parameters continuously acquired by a flexible ultrasound patch; 2. This invention can extract dynamic change trends, abnormal fluctuation amplitudes, and abnormal durations from continuous monitoring data, thereby improving the clinical utilization value of multi-parameter data; 3. This invention incorporates individual baseline conditions, which can reduce the impact of differences in baseline blood flow status among different patients on diagnostic results; 4. This invention eliminates low-quality data segments through contact quality evaluation, which can reduce the impact of poor patch contact, changes in body position, and motion interference on diagnostic results; 5. This invention can output a stroke auxiliary diagnostic score based on the combination relationship between abnormal blood flow velocity, abnormal vascular resistance, abnormal blood flow asymmetry, abnormal blood pressure, and abnormal heart rhythm. 6. This invention can display the source of the main abnormal parameters, making the diagnostic results more interpretable and facilitating clinicians to determine the cause of the abnormality; 7. This invention is applicable to continuous hemodynamic monitoring in emergency departments, neurology departments, ICUs, perioperative periods, and high-risk stroke patients; 8. This invention can provide objective, continuous, and non-invasive monitoring data for early identification, dynamic observation, and auxiliary diagnosis of stroke.

[0151] Example 2 A schematic diagram of the structure of the self-calibration device for the acquisition parameters of the flexible ultrasonic patch provided in Embodiment 2 of this application. Figure 6 This is a schematic diagram of the structure of the self-calibration device for the acquisition parameters of the flexible ultrasonic patch provided in this application embodiment, as shown below. Figure 6 As shown, the device includes: The data acquisition module 610 is used to acquire the ultrasound echo signal, blood flow spectrum signal and cross-sectional signal of the target blood vessel continuously acquired by the flexible ultrasound patch, and to acquire blood pressure parameters and heart rate parameters simultaneously. The contact quality evaluation module 620 is used to evaluate the contact quality of the acquired data based on the ultrasound echo signal, blood flow spectrum signal and vascular cross-sectional signal, and to mark low-quality data segments. The time synchronization module 630 is used to uniformly time-mark blood flow parameters, vascular cross-sectional parameters, blood pressure parameters and heart rate parameters that have passed the quality evaluation, and to perform data matching and resampling according to a preset time window to achieve time synchronization of multi-source parameters; The baseline state establishment module 640 is used to establish an individual baseline state based on historical data obtained through quality assessment when the subject is in a relatively stable state. The individual baseline state includes baseline blood flow velocity, baseline vascular cross-sectional area, baseline blood pressure, and baseline heart rate. The feature construction module 650 is used to compare real-time monitoring data with the corresponding individual basic state, calculate the relative change magnitude and dynamic deviation of each parameter, and generate multi-parameter time series features; The self-calibration module 660 is used to self-calibrate the real-time monitoring parameters based on the multi-parameter time-series characteristics and data quality evaluation results. The self-calibration includes: reducing the weight of low-quality data segments based on the contact quality evaluation results, and calculating the stroke auxiliary diagnosis score based on the multi-parameter time-series characteristics. The output module 670 is used to output the calibrated parameters and the stroke auxiliary diagnostic score.

[0152] The flexible ultrasonic patch acquisition parameter self-calibration device provided in this application embodiment can realize the various processes and corresponding beneficial effects of the above embodiments, and will not be described again here to avoid repetition.

[0153] Example 3 This application also provides a computer device, including a processor and a memory, wherein the memory stores a computer program, and the processor executes the computer program to implement the method described above.

[0154] It should be noted that the computer device in this application embodiment can execute the self-calibration method for the acquisition parameters of the mobile flexible ultrasonic patch as described above, and has corresponding beneficial effects.

[0155] Example 4 This application also provides a readable storage medium storing a program or instructions. When the program or instructions are executed by a processor, they implement the various processes of the above-described self-calibration method for the acquisition parameters of the flexible ultrasonic patch and achieve the same technical effect. To avoid repetition, they will not be described again here.

[0156] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element. Furthermore, it should be noted that the scope of the methods and systems in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0157] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a computer software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0158] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above, which are merely illustrative and not restrictive. Those skilled in the art, under the guidance of this application, can make many modifications without departing from the spirit and scope of the claims, all of which fall within the protection scope of this application.

[0159] The above description is merely a preferred embodiment and the technical principles employed in this application. This application is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions that can be made by those skilled in the art will not depart from the scope of protection of this application. Therefore, although this application has been described in detail through the above embodiments, this application is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of this application, the scope of which is determined by the scope of the claims.

Claims

1. A method for self-calibrating acquisition parameters of a flexible ultrasonic patch, characterized in that, The method includes: The system acquires the ultrasound echo signal, blood flow spectrum signal, and cross-sectional signal of the target blood vessel continuously collected by the flexible ultrasound patch, and simultaneously acquires blood pressure and heart rate parameters. Based on the ultrasound echo signal, blood flow spectrum signal, and vascular cross-sectional signal, the contact quality of the acquired data is evaluated, and low-quality data segments are marked. The blood flow parameters, vascular cross-sectional parameters, blood pressure parameters, and heart rate parameters that have passed the quality evaluation are uniformly time-stamped, and data matching and resampling are performed according to a preset time window to achieve time synchronization of multi-source parameters; Under relatively stable conditions, individual baseline status is established based on historical data obtained through quality assessment. The individual baseline status includes baseline blood flow velocity, baseline vascular cross-sectional area, baseline blood pressure, and baseline heart rate. The real-time monitoring data is compared with the corresponding individual baseline status, and the relative change amplitude and dynamic deviation of each parameter are calculated to generate multi-parameter time series features. Based on the multi-parameter time-series characteristics and data quality evaluation results, the real-time monitoring parameters are self-calibrated. The self-calibration includes: reducing the weight of low-quality data segments based on the contact quality evaluation results, and calculating the stroke auxiliary diagnosis score based on the multi-parameter time-series characteristics. Output the calibrated parameters and the stroke auxiliary diagnostic score.

2. The method according to claim 1, characterized in that, The contact quality evaluation of the collected data includes: At least one of the following was extracted as a quality evaluation index: ultrasound echo intensity, blood flow spectrum continuity, vascular boundary clarity, vascular center stability, patch contact pressure, patch local deformation, and signal-to-noise ratio. The original signal is graded according to the quality evaluation index, and low-quality data segments are marked.

3. The method according to claim 2, characterized in that, The marker of low-quality data segments includes: When any quality evaluation indicator is lower than the corresponding preset standard, the data for the corresponding time period is determined to be a low-quality data segment. Low-quality data segments are excluded from subsequent time synchronization, feature construction, and self-calibration steps, and data quality alerts are generated.

4. The method according to claim 1, characterized in that, The process of uniformly time-stamping blood flow parameters, vascular cross-sectional parameters, blood pressure parameters, and heart rate parameters that have passed quality evaluation, and performing data matching and resampling according to a preset time window, includes: Give each parameter a unified timestamp and map parameters from different sources to the same timeline; Multi-source parameters are statistically analyzed using a sliding window according to a preset time window, where the time window includes 30 seconds, 1 minute, 5 minutes, 10 minutes, 15 minutes, or 30 minutes. Accordingly, the resampling includes: For parameters with different sampling frequencies, time alignment is performed using resampling, interpolation, moving average, or window statistics methods to ensure that each parameter has the same number of sampling points within the same time window.

5. The method according to claim 1, characterized in that, The process of establishing an individual's baseline state based on historical data obtained through quality assessment, under relatively stable conditions for the subjects, includes: During the preset initialization period after system startup, multi-source parameters are collected from subjects in a resting or stable state. Data that passes the quality evaluation are selected as historical data, and statistical values ​​of the historical data are calculated as the basic state of an individual. The statistical values ​​include the mean, median, or mode. The baseline blood flow velocity includes baseline values ​​for peak systolic flow velocity, baseline values ​​for end-diastolic flow velocity, or baseline values ​​for mean blood flow velocity; the baseline vascular cross-sectional area includes baseline values ​​for vascular diameter or vascular cross-sectional area; the baseline blood pressure includes baseline values ​​for systolic blood pressure and diastolic blood pressure; and the baseline heart rate includes baseline values ​​for resting heart rate.

6. The method according to claim 1, characterized in that, The generation of multi-parameter temporal features includes: Extract at least one of the following parameters: mean, maximum, minimum, rate of change, fluctuation amplitude, coefficient of variation, duration of abnormality, mutation index, left and right blood flow asymmetry index, and degree of deviation relative to the individual's baseline state; The multi-parameter time-series features specifically include: The magnitude of decrease in blood flow velocity, the blood flow velocity mutation index, the magnitude of increase in resistance index, the magnitude of increase in pulsatility index, the rate of change in vascular cross-sectional area, the magnitude of blood pressure fluctuation, and abnormal fluctuations in heart rate.

7. The method according to claim 1, characterized in that, The calculation of the stroke auxiliary diagnostic score based on the multi-parameter temporal features includes: Convert the degree of anomaly of each type of parameter into a standardized score; The parameters within each time window are assigned different weights based on the data quality score, with the time window with a high data quality score being assigned a higher weight and the time window with a low data quality score being assigned a lower weight or zero weight. or, A comprehensive abnormality score is calculated based on the combined relationships between abnormal blood flow velocity, abnormal vascular resistance, abnormal blood pressure, and abnormal heart rate. When multiple parameters simultaneously exhibit abnormalities within the same time window, and the duration of the abnormality exceeds a preset threshold, the stroke auxiliary diagnostic score is improved. Output diagnostic results, which include low risk, suspicious abnormality, highly suspicious abnormality, or positive auxiliary diagnostic indications.

8. A self-calibration device for acquisition parameters of a flexible ultrasonic patch, characterized in that, include: The data acquisition module is used to acquire the ultrasound echo signal, blood flow spectrum signal and cross-sectional signal of the target blood vessel continuously acquired by the flexible ultrasound patch, and to acquire blood pressure parameters and heart rate parameters simultaneously. The contact quality evaluation module is used to evaluate the contact quality of the acquired data based on the ultrasound echo signal, blood flow spectrum signal and vascular cross-sectional signal, and to mark low-quality data segments. The time synchronization module is used to uniformly time-mark blood flow parameters, vascular cross-sectional parameters, blood pressure parameters, and heart rate parameters that have passed the quality evaluation, and to perform data matching and resampling according to a preset time window to achieve time synchronization of multi-source parameters. The baseline state establishment module is used to establish an individual baseline state based on historical data obtained through quality assessment when the subject is in a relatively stable state. The individual baseline state includes baseline blood flow velocity, baseline vascular cross-sectional area, baseline blood pressure, and baseline heart rate. The feature construction module is used to compare real-time monitoring data with the corresponding individual baseline status, calculate the relative change magnitude and dynamic deviation of each parameter, and generate multi-parameter time series features; The self-calibration module is used to self-calibrate the real-time monitoring parameters based on the multi-parameter time-series characteristics and data quality evaluation results. The self-calibration includes: reducing the weight of low-quality data segments based on the contact quality evaluation results, and calculating the stroke auxiliary diagnosis score based on the multi-parameter time-series characteristics. The output module is used to output the calibrated parameters and the stroke auxiliary diagnostic score.

9. A computer device, characterized in that, It includes a processor and a memory, the memory storing a computer program, and the processor executing the computer program to implement the method of any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, It stores a computer program thereon, which, when executed by a processor, implements the method of any one of claims 1-7.