Dynamic feedback correction blood glucose real-time monitoring method and system and storage medium

Abstract

The invention relates to the technical field of blood glucose monitoring, and discloses a dynamic feedback correction blood glucose real-time monitoring method and system and a storage medium, and the method comprises the steps: obtaining intertissue fluid data and a venous blood reference value, and determining a sensor drift deviation based on a real-time difference; correcting interstitial fluid data based on the drift deviation and calculating a blood glucose fluctuation rate; if the fluctuation rate exceeds a preset threshold value, adjusting the feedback interval to a high-frequency mode; executing calibration correction in a high-frequency mode, fusing the venous blood reference value to correct interstitial fluid data, and generating a corrected blood glucose monitoring result; evaluating the resource consumption level based on the correction result, if there is an optimization demand, adjusting the feedback interval to a target interval value in a stationary mode, and determining the target interval value as a final feedback rhythm; monitoring parameters are updated based on the final feedback rhythm, so that the system adapts to individual differences and dynamic changes. The blood glucose monitoring accuracy, the system operation efficiency and the individualized adaptive capacity are improved.

Description

Technical Field

[0001] This application relates to the field of blood glucose monitoring technology, and in particular to a method, system and storage medium for real-time blood glucose monitoring with dynamic feedback correction. Background Technology

[0002] Continuous glucose monitoring (CGM) technology, as an important tool for diabetes management, uses implanted sensors to monitor glucose concentration in interstitial fluid in real time. This provides patients with continuous information on blood glucose trends, helping to promptly detect hyperglycemia and hypoglycemia events and guide insulin dosage adjustments and lifestyle interventions. Current technologies, to improve monitoring accuracy, typically use periodic venous or finger-prick blood samples for sensor calibration, and output blood glucose values ​​through fixed-frequency data acquisition and processing, which to some extent meets the basic needs of clinical monitoring.

[0003] However, existing methods still face many challenges in practical applications. First, during long-term implantation, factors such as enzyme activity decay and changes in the local microenvironment cause cumulative drift bias in the sensor, leading to measurements gradually deviating from the true blood glucose concentration. Traditional periodic calibration methods struggle to capture and correct this dynamic drift in real time. Second, significant individual physiological differences exist among patients. For example, the time delay between interstitial fluid glucose and blood glucose varies, and the blood glucose fluctuation characteristics of the same patient differ under different physiological states. Fixed-parameter monitoring modes are ill-suited to these complex and varied individual needs. Furthermore, while high-frequency data acquisition and processing can improve monitoring accuracy, it significantly increases system power consumption and shortens sensor lifespan. Simple threshold triggering or fixed-frequency modes cannot achieve a dynamic balance between monitoring accuracy and resource consumption.

[0004] Therefore, how to accurately extract blood glucose fluctuation characteristics based on real-time identification and correction of sensor drift deviation, adaptively adjust the monitoring frequency according to fluctuation risk and resource consumption status, and achieve personalized parameter optimization through continuous learning of individual physiological characteristics has become a technical challenge that urgently needs to be solved in the field of continuous blood glucose monitoring. Summary of the Invention

[0005] To address the aforementioned technical problems, this application provides a dynamic feedback correction method, system, and storage medium for real-time blood glucose monitoring. This addresses the issues of inaccurate sensor drift correction, delayed capture of blood glucose fluctuation rates, and the inability to dynamically balance monitoring accuracy and resource consumption in existing continuous blood glucose monitoring technologies. It enables adaptive monitoring frequency adjustment and individualized optimization of system parameters based on dual feedback of physiological state and resource consumption.

[0006] In a first aspect, this application provides a dynamic feedback correction method for real-time blood glucose monitoring, the method comprising:

[0007] Step S1: Acquire interstitial fluid data and venous blood reference values, and determine the current sensor drift bias based on the real-time difference between the two.

[0008] Step S2: Correct the interstitial fluid data based on the drift deviation of the sensor, and calculate the blood glucose fluctuation rate based on the corrected interstitial fluid data;

[0009] Step S3: Compare the blood glucose fluctuation rate with a preset fluctuation threshold. If the blood glucose fluctuation rate exceeds the preset fluctuation threshold, adjust the blood glucose monitoring feedback interval to high frequency mode.

[0010] Step S4: In the high-frequency mode, calibration correction is performed based on the adjusted feedback interval, and the interstitial fluid data is corrected by fusing the venous blood reference value to generate a corrected blood glucose monitoring result;

[0011] Step S5: Based on the corrected blood glucose monitoring results, assess the resource consumption level of the blood glucose monitoring system. If there is a need to optimize resource consumption, adjust the feedback interval of blood glucose monitoring to the target interval value in the stable mode and determine it as the final feedback rhythm of the current monitoring cycle.

[0012] Step S6: Update the monitoring parameters of the blood glucose monitoring system based on the final feedback rhythm, so that the blood glucose monitoring system can adapt to individual physiological differences and dynamic changes in blood glucose.

[0013] Secondly, this application provides a dynamic feedback correction real-time blood glucose monitoring system, the system comprising:

[0014] The deviation determination unit is used to acquire interstitial fluid data and venous blood reference values, and determine the current sensor drift deviation based on the real-time difference between the two.

[0015] A rate calculation unit is used to correct the interstitial fluid data based on the drift deviation of the sensor, and to calculate the blood glucose fluctuation rate based on the corrected interstitial fluid data.

[0016] The mode triggering unit is used to compare the blood glucose fluctuation rate with a preset fluctuation threshold. If the blood glucose fluctuation rate exceeds the preset fluctuation threshold, the blood glucose monitoring feedback interval is adjusted to a high-frequency mode.

[0017] The result generation unit is used to perform calibration correction based on the adjusted feedback interval in the high-frequency mode, and to correct the interstitial fluid data by fusing the venous blood reference value, thereby generating a corrected blood glucose monitoring result.

[0018] The resource adjustment unit is used to evaluate the resource consumption level of the blood glucose monitoring system based on the corrected blood glucose monitoring results. If there is a need to optimize resource consumption, the feedback interval of blood glucose monitoring is adjusted to the target interval value in the stable mode and determined as the final feedback rhythm of the current monitoring cycle.

[0019] The parameter adaptation unit is used to update the monitoring parameters of the blood glucose monitoring system based on the final feedback rhythm, so that the blood glucose monitoring system can adapt to individual physiological differences and dynamic changes in blood glucose.

[0020] A third aspect of this application provides a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the aforementioned dynamic feedback correction method for real-time blood glucose monitoring.

[0021] Compared with the prior art, the beneficial effects of the present invention are at least as follows:

[0022] This application achieves a dynamic balance between monitoring accuracy and system energy consumption by constructing a dual-feedback closed-loop control architecture based on blood glucose fluctuation rate and resource consumption level. When blood glucose fluctuates rapidly, it can adaptively switch to high-frequency mode and dynamically optimize the acquisition interval to ensure that key physiological changes are captured in a timely manner. When blood glucose is stable and resource consumption is too high, it automatically reverts to stable mode, significantly reducing the frequency of data acquisition and processing. This maximizes the device's battery life while ensuring monitoring reliability and optimizes the overall utilization efficiency of system resources.

[0023] This application effectively eliminates the interference of physiological delay and device drift by real-time identification and correction of sensor drift deviation, combined with the optimal fusion of interstitial fluid data and venous blood reference values ​​using a Kalman filter. This significantly improves the accuracy and robustness of blood glucose monitoring results. At the same time, it calculates the rate of blood glucose fluctuation based on the corrected data and introduces an abnormal fluctuation state marking mechanism, which can more sensitively capture dangerous events such as hypoglycemia or hyperglycemia, providing more timely and reliable early warning information for clinical intervention.

[0024] This application employs a two-tiered parameter optimization mechanism that combines long-term personalized customization with short-term dynamic fine-tuning. This enables the system to continuously learn the user's individual physiological characteristics and blood glucose variation patterns, and automatically verify and solidify the adapted parameters at the end of each monitoring cycle. This achieves the evolution of the monitoring strategy from a general preset to an individualized one. As usage time increases, the system's adaptability to individual differences continuously improves, further enhancing the stability and accuracy of long-term monitoring and providing users with a truly intelligent continuous blood glucose monitoring experience. Attached Figure Description

[0025] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 This is a flowchart illustrating the steps of a dynamic feedback correction method for real-time blood glucose monitoring in an embodiment of this application.

[0027] Figure 2 This is a schematic diagram illustrating the effect of moving average smoothing filtering in an embodiment of this application;

[0028] Figure 3 This is a schematic diagram illustrating the resource consumption optimization requirement determination in the embodiments of this application;

[0029] Figure 4 This is a structural diagram of a dynamic feedback correction real-time blood glucose monitoring system according to an embodiment of this application. Detailed Implementation

[0030] This application provides a method, system, and storage medium for real-time blood glucose monitoring with dynamic feedback correction. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0031] Example 1:

[0032] For ease of understanding, the specific process of the embodiments of this application is described below, such as... Figure 1 The embodiment of this application shows a real-time blood glucose monitoring method with dynamic feedback correction, which includes:

[0033] Step S1: Acquire interstitial fluid data and venous blood reference values, and determine the current sensor drift deviation based on the real-time difference between the two.

[0034] The determination of the current sensor drift bias includes: acquiring interstitial fluid data through real-time acquisition equipment, and simultaneously acquiring venous blood reference values ​​at the same time point as benchmark data for blood glucose monitoring; comparing the interstitial fluid data and venous blood reference values ​​point by point, calculating the difference between the two, determining the degree of current sensor drift based on the difference sequence, and generating a quantified sensor drift bias value; performing time series analysis on the sensor drift bias to verify the stability characteristics of the sensor drift bias; if the stability index of the sensor drift bias meets the preset conditions, then the sensor drift bias is used as the basis parameter for subsequent blood glucose fluctuation rate correction calculation, and the trend of sensor drift bias over time is recorded to provide a reference for subsequent blood glucose monitoring calibration operations.

[0035] Specifically, in order to address the problem of systematic deviation, or drift bias, between measured values ​​and true blood glucose concentration caused by changes in sensor characteristics such as aging and environmental interference in continuous blood glucose monitoring, this bias is precisely quantified to lay the foundation for accurate calculation and feedback correction of subsequent blood glucose fluctuation rates.

[0036] Specifically, in one embodiment, the current sensor drift deviation is determined as follows: First, glucose concentration data in interstitial fluid is collected in real time using an implanted sensor, serving as the raw data to be corrected; simultaneously, glucose concentration values ​​of venous blood samples are obtained using a fingertip blood glucose meter at the same time, and these values ​​are considered the current true blood glucose reference benchmark; the interstitial fluid data collected at the same time point is compared point by point with the venous blood reference value, and the difference between the two is calculated to obtain a difference sequence that changes over time. This difference sequence directly reflects the sensor's measurement performance in the interstitial fluid environment. Measurement deviation degree: When the difference is positive, it indicates that the sensor reading is higher than the true value, i.e., overestimation; when the difference is 0, it indicates that the sensor reading has not shifted; conversely, it indicates underestimation. Based on this, to quantify the instantaneous rate and trend of drift, a linear regression method can be applied to the difference sequence, such as fitting the relationship between the difference and time using the least squares method, and calculating the slope of the regression line. This slope is the current drift rate of the sensor, i.e., the measurement deviation degree of the sensor, and is used as the quantified value of the drift deviation. In this way, the dynamic changes of sensor drift can be identified in real time, avoiding the long-term accumulation of errors.

[0037] Furthermore, considering the physiological differences among patients, such as the slower interstitial fluid flow in elderly patients potentially leading to larger differences, individualized weighting factors can be introduced to correct the linear regression results when generating drift bias values. This ensures that the final drift bias more closely reflects the patient's actual situation, improving the effectiveness of personalized monitoring. Subsequently, to ensure that the determined drift bias has stable statistical characteristics and can be reliably used for subsequent calculations, time series analysis of the drift bias sequence is necessary to verify its stability. For example, the autocorrelation function can be used to calculate the autocorrelation coefficient of the drift bias sequence. The closer the autocorrelation coefficient is to 1, the stronger the autocorrelation of the sequence over time and the higher its stability. Conversely, a low coefficient may indicate random fluctuations or periodic changes. During the verification process, if the drift bias sequence exhibits significant periodicity, confirming its stability through autocorrelation analysis can avoid using highly volatile and invalid data for subsequent calculations, thereby enhancing the robustness of the system. In addition, for certain dynamic scenarios such as post-exercise blood flow... Significant fluctuations in blood glucose levels lead to trend changes in the deviation sequence. Firstly, the deviation sequence can be subjected to first-order differencing to remove the trend term. Then, an autocorrelation function can be applied to the differencing sequence to more accurately assess its inherent stability. Alternatively, in scenarios such as nighttime monitoring, a sliding window method can be used, for example, calculating the local autocorrelation coefficient within the window hourly to achieve real-time stability verification. When the stability index of the drift deviation meets preset conditions, such as the autocorrelation coefficient exceeding an empirical threshold, the drift deviation can be used as the basis parameter for subsequent blood glucose fluctuation rate correction calculations and stored in a historical database. Simultaneously, its trend over time is recorded. This trend data can be used to guide the frequency of subsequent calibration operations or adjust the parameters of correction modules such as the Kalman filter. For example, when the drift trend is significantly aggravated, the calibration interval can be appropriately shortened to maintain monitoring accuracy. Through these technical means, this solution achieves accurate identification, quantitative verification, and dynamic recording of sensor drift deviation, providing a reliable data foundation for the entire feedback correction process.

[0038] It should be noted that when comparing interstitial fluid data with venous blood reference values, those skilled in the art should understand that there is an inherent physiological time delay between the two, i.e., it takes a certain amount of time for glucose to diffuse from capillaries into the interstitial fluid, typically 5-15 minutes. Therefore, directly subtracting the two types of data collected at the same time does not yield a difference that is entirely equivalent to sensor drift, but rather a result of the combined effect of sensor drift and physiological delay. To address this issue, in this embodiment, before calculating the difference, a physiological delay compensation model is first applied to the interstitial fluid data. This model describes the diffusion process of glucose from blood to interstitial fluid based on a first-order kinetic equation. It estimates the "equivalent blood glucose time point" corresponding to the current interstitial fluid data through deconvolution operations, and then compares it with the corresponding venous blood reference value, thereby more accurately separating the physiological delay and sensor drift.

[0039] Step S2: Correct the interstitial fluid data based on the sensor drift bias, and calculate the blood glucose fluctuation rate based on the corrected interstitial fluid data.

[0040] The calculation of blood glucose fluctuation rate includes: acquiring interstitial fluid data at continuous time points, constructing an interstitial fluid time series dataset, applying a difference method to the interstitial fluid time series dataset to calculate the original change value of interstitial fluid data between adjacent time points; correcting the original change value using sensor drift bias to obtain the corrected change value, determining the quantified value of blood glucose fluctuation rate based on the corrected change value, performing smoothing filtering on the quantified value of blood glucose fluctuation rate, and if the fluctuation amplitude of blood glucose fluctuation rate exceeds a preset range, marking the current blood glucose fluctuation state as an abnormal fluctuation state, recording the historical data sequence of blood glucose fluctuation rate to provide data support for subsequent dynamic adjustment of blood glucose fluctuation rate threshold.

[0041] Specifically, in order to address the problem of how to accurately extract the blood glucose fluctuation rate that reflects the true physiological state from interstitial fluid data carrying sensor drift noise, thereby providing a reliable criterion for subsequent high-frequency mode triggering, in one embodiment, the blood glucose fluctuation rate is calculated in the following manner.

[0042] First, interstitial fluid glucose concentration data are continuously collected at fixed intervals and a time-series dataset is constructed in chronological order. To capture the dynamic trend of blood glucose changes, a difference method is applied to the time-series dataset to calculate the difference between interstitial fluid data points between adjacent time points, obtaining a sequence of raw change values ​​reflecting the absolute change in blood glucose concentration. The raw change values ​​are divided by the time interval between adjacent time points to obtain an uncorrected raw blood glucose fluctuation rate sequence, which contains systematic errors introduced by sensor drift. Therefore, the raw fluctuation rate is corrected using a pre-determined sensor drift deviation. Sensor drift deviation refers to the rate at which sensor measurements deviate systematically over time, obtained through the comparative analysis of interstitial fluid data and venous blood reference values ​​in the above steps, and has the same physical dimensions as the fluctuation rate. The correction process involves subtracting this drift deviation from the raw fluctuation rate to obtain the corrected true blood glucose fluctuation rate, thereby eliminating the influence of sensor drift on the judgment of blood glucose change trends. Based on this, the average value of the corrected fluctuation rate within a preset time period is taken as the quantified value of the blood glucose fluctuation rate at the current moment. This average value can comprehensively reflect the blood glucose change trend over a period of time and reduce the influence of single-point noise.

[0043] To suppress the interference of inherent electronic and physiological noise during the measurement process on the fluctuation rate, the calculated blood glucose fluctuation rate sequence is smoothed and filtered, for example... Figure 2The diagram illustrating the effect of moving average smoothing filtering demonstrates that a moving average filter is used to take the arithmetic mean of the rate values ​​at the current moment and several adjacent moments before and after it, as the smoothed blood glucose fluctuation rate at the current moment. This smoothing process filters out high-frequency noise components, making the fluctuation rate curve smoother and more stable, facilitating subsequent threshold determination. Based on this, the characteristics of blood glucose fluctuation rate changes are monitored in real time, and its fluctuation amplitude is assessed to determine whether it exceeds a preset normal range. The fluctuation amplitude can be quantified by calculating the standard deviation of the rate sequence within a preset time window; a larger standard deviation indicates more severe blood glucose fluctuations. If the fluctuation amplitude within the current window exceeds a preset threshold, the current blood glucose fluctuation state is determined to be an abnormal fluctuation state and marked. This marking can be used to trigger an early warning mechanism or as a basis for subsequent analysis. Simultaneously, historical data sequences of blood glucose fluctuation rates are continuously recorded, including the rate value at each moment and its corresponding timestamp, as well as the marking information of abnormal fluctuation states. This historical data is stored in a local database for subsequent analysis of individual blood glucose fluctuation patterns and provides data support for the dynamic adjustment of blood glucose fluctuation rate thresholds. For example, the preset fluctuation threshold can be dynamically optimized based on the statistical distribution of historical rates to better suit the individual physiological characteristics of users and avoid misjudgments or omissions caused by fixed thresholds in different individuals or under different physiological states. Through the above technical means, this technical solution achieves accurate extraction of the real blood glucose fluctuation rate from noisy data and performs smoothing, anomaly identification, and historical recording, providing a reliable data foundation for subsequent adjustment of the feedback interval based on fluctuation rate.

[0044] Step S3: Compare the blood glucose fluctuation rate with the preset fluctuation threshold. If the blood glucose fluctuation rate exceeds the preset fluctuation threshold, adjust the blood glucose monitoring feedback interval to high frequency mode.

[0045] The adjustment of the blood glucose monitoring feedback interval to high-frequency mode includes: acquiring the current blood glucose fluctuation rate value in real time and comparing it with a preset fluctuation threshold; if the blood glucose fluctuation rate exceeds the preset fluctuation threshold, triggering the blood glucose monitoring system to switch to high-frequency feedback monitoring mode; calculating the feedback interval duration adapted to the current blood glucose fluctuation state through an adaptive adjustment algorithm; shortening the interstitial fluid data acquisition cycle and blood glucose data processing cycle based on the calculated feedback interval duration; continuously monitoring the changing trend of blood glucose fluctuation rate; dynamically adjusting the feedback interval in real time; and recording the current feedback interval adjustment parameters when the blood glucose fluctuation rate recovers to below the preset fluctuation threshold as a reference for subsequent mode switching.

[0046] Specifically, to address the issue of how to adaptively adjust the monitoring frequency when blood glucose fluctuates rapidly to ensure that key physiological changes are captured in a timely manner, while avoiding resource waste caused by a fixed high-frequency mode, in one embodiment, the feedback interval of blood glucose monitoring is adjusted to a high-frequency mode in the following way.

[0047] First, the current blood glucose fluctuation rate is acquired in real time. This fluctuation rate is calculated based on the interstitial fluid data collected at the current monitoring frequency in the steps described above, reflecting the true range of blood glucose concentration changes per unit time. The current fluctuation rate is then compared with a preset fluctuation threshold. The preset fluctuation threshold is a critical value pre-set according to clinical safety requirements and individual physiological characteristics, used to determine whether the current blood glucose change is in a rapid fluctuation state that requires enhanced monitoring. If the current fluctuation rate exceeds the preset threshold, it is determined that blood glucose is in a rapid change phase, and the monitoring density needs to be increased to capture detailed changes. At this time, the system is switched to high-frequency feedback monitoring mode. High-frequency feedback monitoring mode is a working state with a significantly shorter data acquisition and processing cycle than the conventional mode, aiming to achieve close tracking of rapid changes in blood glucose.

[0048] After the high-frequency mode is triggered, more dense interstitial fluid data is immediately collected at a preset initial high-frequency interval, such as half of the regular interval. After obtaining the first batch of high-frequency data points, the blood glucose fluctuation rate at the current moment is recalculated based on these newly collected data. This recalculated fluctuation rate can more accurately reflect the true details of blood glucose changes. Subsequently, an adaptive adjustment algorithm is used to calculate the optimal feedback interval duration that adapts to the current blood glucose fluctuation state. As a preferred implementation, the adaptive adjustment algorithm can adopt a proportional-integral-derivative (PID) control algorithm, which is a classic control method that dynamically adjusts based on feedback deviation. The principle of this method is to use the weighted sum of the proportional term responding to the magnitude of the current deviation, the integral term accumulating the effect of historical deviations, and the differential term predicting the future trend of deviation changes, as the control output. In this embodiment, the input of the algorithm is the deviation value between the recalculated blood glucose fluctuation rate and the preset fluctuation threshold, and the output is the adjustment coefficient of the feedback interval. Multiplying this adjustment coefficient by the currently used monitoring interval yields the optimized feedback interval duration. In this way, the monitoring frequency can be dynamically optimized based on the fluctuation details revealed by high-frequency data. The more intense the fluctuation, the shorter the monitoring interval, thereby achieving precise allocation of monitoring resources to high-risk periods.

[0049] Based on the optimized feedback interval duration, the acquisition cycle of interstitial fluid data and the processing cycle of blood glucose data are adjusted accordingly to ensure that the subsequent data acquisition and processing frequency maintains optimal matching with the current fluctuation state. On this basis, the system continues to operate in the optimized high-frequency mode, and the fluctuation rate is periodically recalculated based on the latest acquired data to dynamically adjust the feedback interval in real time, forming an online closed-loop optimization process. For example, if the fluctuation rate continues to increase, the feedback interval can be further shortened; conversely, if the fluctuation rate tends to level off, the interval can be appropriately extended to avoid oversampling. This dynamic adjustment mechanism ensures that the monitoring frequency is always synchronized with the real-time needs of blood glucose changes. When the blood glucose fluctuation rate gradually recovers to below the preset fluctuation threshold, it indicates that blood glucose has returned to a relatively stable state. At this time, the feedback interval adjustment parameters finally adopted in the current high-frequency mode are recorded, including the optimized interval duration and the correlation coefficient of the adaptive adjustment algorithm, and these parameters are stored in the database as historical reference data. These records can be used for rapid response in subsequent similar fluctuation scenarios. When the high-frequency mode is triggered again, the historical parameters can be directly called as the initial settings, thereby accelerating the response process and reducing the overhead of repeated calculations. Through the aforementioned technical means, this technical solution achieves adaptive high-frequency monitoring based on blood glucose fluctuation rate. While ensuring that key physiological changes are captured in a timely manner, it optimizes resource utilization efficiency through dynamic adjustment and parameter recording mechanisms, and provides flexible rhythm control for the entire feedback correction process.

[0050] Step S4: In high-frequency mode, perform calibration correction based on the adjusted feedback interval, and correct the interstitial fluid data by fusing venous blood reference values ​​to generate corrected blood glucose monitoring results.

[0051] The process of generating corrected real-time blood glucose monitoring results includes: acquiring the latest interstitial fluid data periodically based on the adjusted high-frequency mode feedback interval; performing data fusion processing on the latest interstitial fluid data and venous blood reference values ​​using a Kalman filter; generating drift-corrected blood glucose monitoring results based on the data fusion processing results; comparing the corrected blood glucose monitoring results with historical blood glucose monitoring data to verify the calibration effect; if the calibration effect meets the preset standard, updating the calibration parameters of the blood glucose monitoring sensor, recording the corrected blood glucose monitoring results, and providing data support for subsequent resource consumption assessment.

[0052] Specifically, in order to address the problem of how to effectively fuse real-time collected interstitial fluid data with sparse but accurate venous blood reference values ​​in high-frequency monitoring mode to generate accurate blood glucose monitoring results after drift correction, in one embodiment, the generation of corrected real-time blood glucose monitoring results is achieved in the following way.

[0053] First, based on the dynamically adjusted feedback interval in high-frequency mode, the sensor is periodically triggered to collect interstitial fluid glucose concentration data at the current moment. The collected analog signal is then converted into a digital signal, which serves as the real-time observation input for the Kalman filter. The Kalman filter is an optimal recursive algorithm used to estimate the state of a dynamic system from noisy observation data. Its core principle is to minimize the covariance of the estimation error through iterative prediction and update steps. In this embodiment, the state vector of the Kalman filter is constructed to include the current true blood glucose estimate and the current sensor drift deviation estimate. The former reflects the physiological state, while the latter characterizes the systematic deviation of the sensor measurement over time. The system state transition model describes the natural evolution of blood glucose values ​​over time and the characteristics of drift deviation changes. The process noise covariance quantifies the uncertainty of the model prediction.

[0054] At each monitoring moment, the filter first performs a prediction step, calculating the current state prediction value and its prediction covariance based on the state estimate and state transition model from the previous moment. Then, it proceeds to an update step: when only interstitial fluid data arrives, this data is treated as a regular observation, with its observation noise covariance preset according to sensor characteristics. The filter calculates the Kalman gain, a function of the prediction covariance and the observation noise covariance, used to balance the reliability of the prediction result with the current observation value. The interstitial fluid data and the state prediction value are then weighted and fused to obtain the updated state estimate. When venous blood reference data is available... When the reference value is reached, it is also used as an input filter for the observation, but its observation noise covariance is set to be much smaller than that of the interstitial fluid data to reflect its high confidence as a true blood glucose benchmark. At this time, the Kalman filter uses the same update mechanism to jointly correct all variables in the state vector, including the blood glucose estimate and the drift bias estimate, using this high-confidence observation. This effectively corrects the sensor's drift characteristics while updating the blood glucose estimate. Through this unified update framework, the filter can continuously track the real changes in blood glucose and achieve system-level calibration when the venous blood reference value is reached.

[0055] Based on the fusion processing results of the filter, the current real blood glucose estimate is extracted from the updated state vector and output as the corrected blood glucose monitoring result after drift correction. To ensure the reliability of the calibration correction, the corrected blood glucose monitoring result at the current moment is compared and verified with historical blood glucose monitoring data. If the deviation is less than the preset calibration quality standard, the calibration effect is determined to be effective. Then, the sensor calibration parameters are updated according to the verification results, such as adjusting the sensor gain factor to better match the sensor characteristics with the current individual physiological state. At the same time, the corrected blood glucose monitoring result at each moment and its corresponding timestamp are stored in a local database as the data basis for subsequent resource consumption assessment and fluctuation rate calculation. Through the above technical means, this technical solution realizes data fusion calibration based on Kalman filtering in high-frequency mode. While making full use of real-time interstitial fluid data, it effectively suppresses sensor drift with the help of high-confidence venous blood reference values, generating accurate and reliable blood glucose monitoring results, and providing high-quality data input for subsequent resource optimization decisions.

[0056] Step S5: Based on the corrected blood glucose monitoring results, assess the resource consumption level of the blood glucose monitoring system. If there is a need to optimize resource consumption, adjust the feedback interval of blood glucose monitoring to the target interval value in the stable mode and determine it as the final feedback rhythm of the current monitoring cycle.

[0057] The assessment of the resource consumption level of the blood glucose monitoring system includes: acquiring the computational load data generated by the blood glucose monitoring system during the generation of corrected blood glucose monitoring results; comparing the computational load data with the historical load records of the blood glucose monitoring system to determine the current resource consumption level of the blood glucose monitoring system; if the current resource consumption level of the blood glucose monitoring system exceeds a preset load threshold, it is determined that the blood glucose monitoring system has a need for resource consumption optimization; performing trend analysis on the time series data of resource consumption level to determine optimization priorities; generating a resource allocation adjustment plan for the blood glucose monitoring system based on the optimization priorities; and recording the change process of the resource consumption level of the blood glucose monitoring system to provide a reference for subsequent optimization decisions of the blood glucose monitoring system.

[0058] Specifically, to address the issue of surging resource consumption caused by intensive data acquisition and complex algorithm calculations in high-frequency monitoring mode, a balance between device power consumption and computing resources is achieved by real-time evaluation of computing load and dynamic optimization of monitoring frequency, while ensuring monitoring accuracy. In one embodiment, the resource consumption level of the blood glucose monitoring system is evaluated in the following way.

[0059] First, during operation, surrogate indicators related to resource consumption are collected in real time, rather than directly measuring the precise load of a single operation. These surrogate indicators include the number of data acquisitions per unit time, the number of Kalman filter updates, and the proportion of high-frequency mode duration to the current monitoring cycle. These indicators can indirectly but effectively reflect the resource usage intensity of the system and are easy to implement in embedded systems using simple counters, with negligible measurement overhead. The above multiple surrogate indicators are then fused through preset weighting coefficients to obtain a quantitative indicator value that can comprehensively characterize the current resource consumption level of the system. This real-time resource consumption indicator provides an operable quantitative basis for subsequent comparative analysis.

[0060] After obtaining the current resource consumption index, it is compared and analyzed with the historical resource consumption records stored in the local database. The historical resource consumption records are the time series of indicators that the system has continuously accumulated over a preset period of time. By calculating the statistical deviation between the current index and the corresponding data in the historical series, it can be determined whether the current load is within the normal fluctuation range or significantly deviates from the historical normal. If the current index deviates significantly from the historical normal, the resource consumption level is divided into normal, high, or overload levels according to the degree of deviation. This comparison method based on historical data can effectively identify resource consumption peaks caused by sudden high-frequency monitoring, thereby improving the system's ability to perceive abnormal load states.

[0061] Based on the current resource consumption level, further determine whether there is a need for resource consumption optimization; such as... Figure 3 The diagram illustrating the resource consumption optimization requirement determination uses a preset resource consumption threshold. This threshold can be a multiple of the historical average consumption or an absolute upper limit of system resource usage, such as the critical point at which the processor cannot complete all tasks within a preset time. If the current resource consumption level exceeds this preset threshold, the system is determined to have a resource optimization requirement, and the subsequent optimization process is triggered. This determination mechanism ensures that optimization is only initiated when resource consumption is truly overloaded, avoiding frequent adjustments due to slight fluctuations, thereby maintaining the system's operational stability.

[0062] Once an optimization need is identified, trend analysis is performed on the time-series data of resource consumption levels to determine the urgency of optimization. Trend analysis involves collecting resource consumption data from several recent monitoring periods to form a time series, and then smoothing the series using a moving average method to eliminate instantaneous noise interference and extract the trend of resource consumption changes. The slope of the smoothed trend can be used to determine whether resource consumption is rising, falling, or stable. Simultaneously, a comprehensive assessment is made in conjunction with the current blood glucose fluctuation rate: if resource consumption is on a sharp upward trend and the blood glucose fluctuation rate also exceeds a preset threshold, it indicates that the system is in a critical state of high load and high risk, at which point the urgency of optimization is highest, and resource consumption reduction measures should be prioritized. If resource consumption is high but the trend is stable, or blood glucose fluctuations have eased, the urgency of optimization is correspondingly reduced. This multi-dimensional assessment method ensures that the system only initiates adjustments to the monitoring frequency when most needed.

[0063] Based on the determined optimization urgency, a corresponding adjustment range for the feedback interval is generated. Specifically, the optimization urgency is mapped to an extension coefficient of the feedback interval; the higher the urgency, the larger the extension coefficient, meaning a greater reduction in monitoring frequency. For example, at the highest urgency, the current feedback interval can be multiplied by a large extension coefficient to significantly reduce sampling density; at lower urgency, a smaller extension coefficient is used for fine-tuning. After generating the adjustment range, simulations can be performed on historical data to estimate the expected reduction in resource consumption after adjustment, confirming that it can effectively reduce resource consumption while ensuring basic monitoring accuracy.

[0064] Finally, the current resource consumption level, the generated adjustment range, and the changes in resource consumption after the adjustment are fully recorded in the log database, and these records are periodically summarized to update the historical consumption distribution. These records not only provide a reference for current optimization decisions, but also accumulate data assets for the continuous optimization of system parameters in long-term operation. Through the above technical means, this technical solution realizes real-time monitoring, dynamic evaluation, and intelligent optimization of system resource consumption. By adjusting the feedback interval as the core control method, the system maximizes the battery life of the device and maintains stable operation while ensuring that the accuracy of blood glucose monitoring is not affected.

[0065] The process of adjusting the blood glucose monitoring feedback interval to the target interval value in stable mode and determining it as the final feedback rhythm for the current monitoring cycle includes: if it is determined that the blood glucose monitoring system has a need for resource consumption optimization, the blood glucose monitoring system is triggered to switch to stable monitoring mode. Through an adaptive adjustment algorithm, the target feedback interval value in stable monitoring mode is calculated. Based on the target feedback interval value, the collection cycle of interstitial fluid data and the processing cycle of blood glucose data are extended. By balancing the monitoring accuracy requirements and resource utilization efficiency, the adjustment range of the feedback interval is optimized. If the resource consumption level of the blood glucose monitoring system returns to the preset normal range after the feedback interval is adjusted, the current feedback interval is determined as the final feedback rhythm, and the relevant parameters of the final feedback rhythm are recorded to provide data basis for subsequent dynamic adjustment of the feedback interval.

[0066] Specifically, to address the issue of how to adaptively adjust the monitoring frequency back to a stable state and determine the optimal feedback interval while ensuring basic monitoring accuracy when the system consumes too much resources due to high-frequency monitoring, in one embodiment, the feedback interval of blood glucose monitoring is adjusted to the target interval value in the stable mode and determined as the final feedback rhythm of the current monitoring cycle through the following method.

[0067] When the system determines that there is a need to optimize resource consumption based on the above steps, it triggers a switch to a stable monitoring mode. The stable monitoring mode is a working state with a longer data acquisition and processing cycle compared to the high-frequency mode. Its goal is to reduce the system load during periods of stable blood glucose fluctuation, thereby extending the device's battery life and freeing up computing resources. After switching to stable mode, a preliminary target feedback interval value is first calculated based on the current blood glucose fluctuation rate and resource consumption level using an adaptive adjustment algorithm. As a preferred implementation, the adaptive adjustment algorithm can adopt a proportional-integral-derivative control algorithm. In this embodiment, the input of the algorithm includes the deviation between the current blood glucose fluctuation rate and the preset fluctuation threshold, and the deviation between the current resource consumption level and the preset load threshold. The algorithm calculates the preliminary feedback interval adjustment amount based on the weighted sum of these two normalized deviations, and then obtains the preliminary target interval value.

[0068] After obtaining the initial target interval, the system does not apply it directly, but instead performs amplitude constraint optimization to ensure that monitoring accuracy is not excessively sacrificed while reducing resource consumption. The amplitude constraint optimization is based on prior knowledge rather than real-time multi-objective optimization: the system presets a maximum allowable interval extension coefficient, which is pre-determined according to sensor characteristics and clinical safety requirements, ensuring that even under the most stable conditions, the sampling frequency will not fall below the lower limit required to guarantee basic monitoring capabilities. At the same time, the system refers to the interval values ​​successfully applied under the same or similar blood glucose fluctuation conditions in the historical records to correct the initial target interval. If the initial target interval exceeds the range of successful applications in the past, it is adjusted back to the maximum value within that range. Through this amplitude constraint based on prior knowledge and historical experience, the system ensures that the adjustment will not lead to unacceptable loss of accuracy even before the actual effect of the adjustment is known, based on a conservative principle.

[0069] Based on the target feedback interval value optimized by amplitude constraints, the system correspondingly extends the data acquisition cycle of interstitial fluid and the processing cycle of blood glucose data to match the frequency of data acquisition and processing with the current stable state. After the adjustment is completed, a period of stable monitoring begins, during which changes in resource consumption levels are continuously monitored and the evolution of blood glucose fluctuation rates is observed simultaneously. If the resource consumption level gradually decreases within a preset time and recovers to a preset normal range, such as being lower than a certain percentage of the historical average load, and no missed blood glucose events or obvious data anomalies occur during this period due to the reduced sampling frequency, it indicates that the current adjustment scheme effectively reduces resource consumption without causing substantial damage to monitoring reliability. At this point, the currently used feedback interval is determined as the final feedback rhythm of the current monitoring cycle. The final feedback rhythm refers to the acquisition and processing frequency that the system will stably execute for the remaining time of this monitoring cycle.

[0070] The relevant parameters of the final feedback rhythm, including interval duration, corresponding blood glucose fluctuation rate range, resource consumption level, and actual effects after adjustment, are fully recorded in a local database as historical reference data. These records can be used for rapid response in similar scenarios in the future: when similar resource consumption states and blood glucose fluctuation characteristics occur again, historical parameters can be directly called as the reference for initial settings, thereby accelerating the adjustment process of the stable mode and reducing redundant calculation overhead. Through the above technical means, this technical solution achieves adaptive callback of monitoring frequency when resources are overloaded, ensures the safety of adjustment through amplitude constraints based on prior knowledge, and continuously optimizes the long-term operating performance of the system through the parameter recording mechanism.

[0071] Step S6: Update the monitoring parameters of the blood glucose monitoring system based on the final feedback rhythm, so that the blood glucose monitoring system can adapt to individual physiological differences and dynamic changes in blood glucose.

[0072] The process of updating the blood glucose monitoring system's parameters to adapt to individual physiological differences and dynamic changes in blood glucose includes: adjusting the system's operating parameters based on the final feedback rhythm to match current blood glucose monitoring needs; collecting and analyzing users' individual physiological difference data; customizing and updating the system's operating parameters; dynamically adjusting the configuration strategy of the operating parameters based on real-time trends in blood glucose changes; verifying the monitoring stability after adjusting the operating parameters; determining the optimization effect of the parameter adjustments based on the verification results; if the optimization effect meets preset requirements, saving the current operating parameters as the system's default configuration; and recording the adjustment history of the operating parameters to provide reference data for subsequent personalized optimization.

[0073] Specifically, to address the issue of how to solidify the monitoring rhythm determined during the aforementioned dynamic optimization process into the baseline parameters for system operation, and to achieve adaptive optimization of the monitoring system across multiple time scales through long-term learning of individual physiological characteristics and short-term response to dynamic changes in blood glucose, in one embodiment, the monitoring parameters of the blood glucose monitoring system are updated to adapt the system to individual physiological differences and dynamic changes in blood glucose in the following way.

[0074] First, the operating parameters are adjusted based on the final feedback rhythm as the baseline configuration for the current monitoring cycle. The final feedback rhythm is the acquisition and processing frequency determined in the previous steps by balancing resource consumption and monitoring accuracy. This rhythm value is set as the trigger cycle parameter of the internal timer to ensure that subsequent data acquisition and calibration correction operations are executed stably at this frequency. At the same time, other operating parameters related to calibration correction operations are updated synchronously, such as the update frequency of the process noise covariance matrix in the Kalman filter, to keep it consistent with the new feedback rhythm, thereby ensuring that each algorithm module operates in a coordinated manner within a unified timing framework. Thus, the baseline parameter configuration for the current monitoring cycle is established.

[0075] Building upon the baseline parameter configuration, the system further constructs a two-layer optimization mechanism: long-term personalized customization and short-term dynamic fine-tuning. These two mechanisms operate on cross-cycle and intra-cycle timescales, respectively. Long-term personalized customization updates the baseline parameters based on individual difference data accumulated by the user during historical monitoring cycles, updating the baseline parameters before the end of the current cycle and the start of the next. Individual difference data includes statistical characteristics of the user's historical blood glucose fluctuation rate, such as the long-term distribution patterns of mean, standard deviation, fluctuation frequency, and sensor drift deviation values, as well as the typical range of resource consumption levels. At the end of each monitoring cycle, a set of personalized parameter adjustment coefficients is recalculated based on all historical data up to the current time. For example, for users with a large standard deviation in blood glucose fluctuation rate, the preset value of the fluctuation rate threshold triggering the high-frequency feedback mode is appropriately increased to avoid misjudging normal physiological fluctuations as abnormal events requiring high-frequency monitoring. These personalized adjustment coefficients are applied to the system's baseline parameters to generate an updated parameter set adapted to the user's physiological characteristics. Subsequently, the parameter verification phase begins: the updated parameter set is applied to historical data playback, simulating its performance over a complete past cycle, and compared with reference values ​​of venous blood collected during the same period to calculate monitoring stability indicators, including consistency correlation coefficient and mean absolute relative error. If the simulation verification results meet the preset requirements, the parameter set is confirmed as the default configuration for the next cycle and written to non-volatile memory. Through this mechanism of "verification at the end of the cycle and effectiveness at the beginning of the cycle", it is ensured that each parameter update is fully verified to avoid the impact of faulty parameters on actual monitoring.

[0076] Short-term dynamic fine-tuning involves temporarily adjusting some operating parameters within the current monitoring period based on the real-time monitored dynamic trend of blood glucose changes, without changing the baseline configuration. As the system operates according to the final feedback rhythm, time-series data is continuously collected, including blood glucose fluctuation rate values ​​and drift deviation values ​​over multiple consecutive periods. Trend analysis is performed on this data; for example, the moving average and moving standard deviation of blood glucose fluctuation rate are calculated using a sliding window method. By identifying the monotonic change trend of the moving average over multiple consecutive windows and combining it with the synchronous change of the moving standard deviation, it is determined whether there is a clear upward or downward trend. When a clear upward or downward trend is identified... When a trend is identified, the configuration of relevant parameters is temporarily fine-tuned accordingly. For example, if a clear upward trend in the rate of blood glucose fluctuation is detected, indicating that blood glucose is about to enter a rapid change phase, the system temporarily lowers the preset value of the fluctuation rate threshold that triggers the high-frequency feedback mode within the current cycle. This allows the system to enter the high-frequency monitoring state earlier to cope with the accelerated changes in blood glucose. This temporary adjustment is only effective within the current cycle and automatically reverts to the baseline configuration at the end of the cycle, without affecting the default parameters for the next cycle. In this way, short-term dynamic fine-tuning achieves a rapid response to real-time physiological changes while maintaining a clear boundary with long-term personalized customization.

[0077] This solution comprehensively records the historical information of every parameter adjustment, including personalized customization records across cycles and dynamic fine-tuning records within cycles. Each record includes a timestamp, adjustment type (cross-cycle / within-cycle), a snapshot of the parameters before adjustment, the reason for the adjustment, a snapshot of the parameters after adjustment, and an evaluation of the effect after adjustment. As users use the system for longer periods, this historical data becomes increasingly rich, providing increasingly accurate references for subsequent personalized optimization. This allows the system to continuously learn and adapt to the physiological characteristics and behavioral patterns of specific users, forming a closed-loop evolutionary process of collaborative optimization at different time scales. Through the above technical means, this solution achieves a leap from single dynamic optimization to long-term adaptive optimization, enabling the blood glucose monitoring system to maintain a balance between high accuracy and high efficiency at different times and under different physiological states.

[0078] Through the coordination of the above steps, this application achieves a dynamic balance between blood glucose monitoring accuracy and resource consumption, significantly improving the accuracy of blood glucose monitoring, system operating efficiency, and individualized adaptive capabilities.

[0079] Example 2:

[0080] The above describes a dynamic feedback correction method for real-time blood glucose monitoring in embodiments of this application. The following describes a dynamic feedback correction system for real-time blood glucose monitoring in embodiments of this application, such as... Figure 4 As shown in the figure, a dynamic feedback correction real-time blood glucose monitoring system according to an embodiment of this application includes:

[0081] The deviation determination unit is used to acquire interstitial fluid data and venous blood reference values, and determine the current sensor drift deviation based on the real-time difference between the two.

[0082] The rate calculation unit is used to correct interstitial fluid data based on sensor drift deviation and to calculate blood glucose fluctuation rate based on the corrected interstitial fluid data.

[0083] The mode triggering unit is used to compare the blood glucose fluctuation rate with a preset fluctuation threshold. If the blood glucose fluctuation rate exceeds the preset fluctuation threshold, the blood glucose monitoring feedback interval is adjusted to a high-frequency mode.

[0084] The result generation unit is used to perform calibration correction based on the adjusted feedback interval in high-frequency mode, and to correct the interstitial fluid data by fusing venous blood reference values ​​to generate corrected blood glucose monitoring results.

[0085] The resource adjustment unit is used to assess the resource consumption level of the blood glucose monitoring system based on the corrected blood glucose monitoring results. If there is a need to optimize resource consumption, the feedback interval of blood glucose monitoring is adjusted to the target interval value in the stable mode and determined as the final feedback rhythm of the current monitoring cycle.

[0086] The parameter adaptation unit is used to update the monitoring parameters of the blood glucose monitoring system based on the final feedback rhythm, so that the blood glucose monitoring system can adapt to individual physiological differences and dynamic changes in blood glucose.

[0087] Through the synergistic collaboration of the aforementioned components, the accuracy of blood glucose monitoring, system operating efficiency, and individualized adaptive capabilities are further improved.

[0088] This application also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when the instructions are executed on a computer, cause the computer to perform the steps of the dynamic feedback correction real-time blood glucose monitoring method.

[0089] In summary, this application achieves a dynamic balance between monitoring accuracy and system energy consumption by constructing a dual-feedback closed-loop control architecture based on blood glucose fluctuation rate and resource consumption level. First, by using real-time comparison of interstitial fluid data and venous blood reference values, sensor drift deviation is accurately identified and quantified, laying the foundation for subsequent calibration. Based on this, the blood glucose fluctuation rate is calculated using the calibrated data, providing a reliable criterion for mode switching. When rapid blood glucose fluctuations are detected, the system adaptively switches to a high-frequency mode and performs precise calibration by fusing reference values ​​using a Kalman filter. When resource consumption is too high, it automatically reverts to a stable mode and determines the optimal feedback interval. Furthermore, through a two-layer parameter optimization mechanism combining long-term personalized customization and short-term dynamic fine-tuning, the system continuously learns individual user characteristics and verifies and solidifies parameters at the end of the cycle, achieving continuous evolution of the monitoring strategy. This application significantly improves the accuracy of blood glucose monitoring, system operating efficiency, and personalized adaptive capabilities, maximizing device endurance while ensuring monitoring reliability, and providing a complete, closed-loop, and evolvable intelligent solution for the field of continuous blood glucose monitoring.

[0090] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the system and unit described above can be referred to the corresponding process in the above method embodiments, and will not be repeated here.

[0091] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0092] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the above embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for real-time blood glucose monitoring with dynamic feedback correction, characterized in that, The method includes: Step S1: Acquire interstitial fluid data and venous blood reference values, and determine the current sensor drift bias based on the real-time difference between the two. Step S2: Correct the interstitial fluid data based on the drift deviation of the sensor, and calculate the blood glucose fluctuation rate based on the corrected interstitial fluid data; Step S3: Compare the blood glucose fluctuation rate with a preset fluctuation threshold. If the blood glucose fluctuation rate exceeds the preset fluctuation threshold, adjust the blood glucose monitoring feedback interval to high frequency mode. Step S4: In the high-frequency mode, calibration correction is performed based on the adjusted feedback interval, and the interstitial fluid data is corrected by fusing the venous blood reference value to generate a corrected blood glucose monitoring result; Step S5: Based on the corrected blood glucose monitoring results, assess the resource consumption level of the blood glucose monitoring system. If there is a need to optimize resource consumption, adjust the feedback interval of blood glucose monitoring to the target interval value in the stable mode and determine it as the final feedback rhythm of the current monitoring cycle. Step S6: Update the monitoring parameters of the blood glucose monitoring system based on the final feedback rhythm, so that the blood glucose monitoring system can adapt to individual physiological differences and dynamic changes in blood glucose.

2. The method for real-time blood glucose monitoring with dynamic feedback correction according to claim 1, characterized in that, Determining the current sensor drift deviation in step S1 includes: Interstitial fluid data is acquired in real time by a data acquisition device, and venous blood reference values ​​at the same time point are acquired as benchmark data for blood glucose monitoring. The interstitial fluid data and venous blood reference values ​​at the same time point are compared point by point, the difference between the two is calculated, and the degree of current sensor drift is determined based on the difference sequence, generating a quantified sensor drift deviation value. Time series analysis is performed on the drift deviation of the sensor to verify its stability characteristics. If the stability index of the drift deviation of the sensor meets the preset conditions, the drift deviation of the sensor is used as the basic parameter for subsequent blood glucose fluctuation rate correction calculation. The change trend of the drift deviation of the sensor over time is recorded to provide a reference for subsequent blood glucose monitoring calibration operations.

3. The method for real-time blood glucose monitoring with dynamic feedback correction according to claim 1, characterized in that, The calculation of blood glucose fluctuation rate in step S2 includes: Acquire interstitial fluid data at consecutive time points, construct an interstitial fluid time series dataset, and apply the difference method to the interstitial fluid time series dataset to calculate the original change value of interstitial fluid data between adjacent time points; The original change value is corrected by using the drift deviation of the sensor to obtain the corrected change value. The quantified value of the blood glucose fluctuation rate is determined based on the corrected change value. The quantified value of the blood glucose fluctuation rate is then smoothed and filtered. If the fluctuation amplitude of the blood glucose fluctuation rate exceeds a preset range, the current blood glucose fluctuation state is marked as an abnormal fluctuation state. The historical data sequence of the blood glucose fluctuation rate is recorded to provide data support for the subsequent dynamic adjustment of the blood glucose fluctuation rate threshold.

4. The method for real-time blood glucose monitoring with dynamic feedback correction according to claim 1, characterized in that, The step S3, adjusting the blood glucose monitoring feedback interval to high-frequency mode, includes: The system acquires the blood glucose fluctuation rate value at the current moment in real time and compares it with the preset fluctuation threshold. If the blood glucose fluctuation rate exceeds the preset fluctuation threshold, the blood glucose monitoring system is triggered to switch to high-frequency feedback monitoring mode. An adaptive adjustment algorithm is used to calculate the feedback interval duration that adapts to the current blood glucose fluctuation state. Based on the calculated feedback interval duration, the collection cycle of interstitial fluid data and the processing cycle of blood glucose data are shortened. The changing trend of the blood glucose fluctuation rate is continuously monitored, and the feedback interval is dynamically adjusted in real time. When the blood glucose fluctuation rate recovers to below the preset fluctuation threshold, the current feedback interval adjustment parameters are recorded and used as a reference for subsequent mode switching.

5. The method for real-time blood glucose monitoring with dynamic feedback correction according to claim 1, characterized in that, The step S4 involves generating the corrected real-time blood glucose monitoring results, including: Based on the adjusted high-frequency mode feedback interval, the latest interstitial fluid data is acquired periodically, and the latest interstitial fluid data and venous blood reference values ​​are fused using a Kalman filter. Based on the data fusion processing results, a drift-corrected blood glucose monitoring result is generated. The corrected blood glucose monitoring result is compared with historical blood glucose monitoring data to verify the calibration effect. If the calibration effect meets the preset standard, the calibration parameters of the blood glucose monitoring sensor are updated, and the corrected blood glucose monitoring result is recorded to provide data support for subsequent resource consumption assessment.

6. The method for real-time blood glucose monitoring with dynamic feedback correction according to claim 1, characterized in that, The assessment of the resource consumption level of the blood glucose monitoring system in step S5 includes: The computational load data generated by the blood glucose monitoring system during the generation of the corrected blood glucose monitoring result is obtained. The computational load data is compared with the historical load records of the blood glucose monitoring system to determine the current resource consumption level of the blood glucose monitoring system. If the current resource consumption level of the blood glucose monitoring system exceeds the preset load threshold, it is determined that the blood glucose monitoring system has a resource consumption optimization requirement. Trend analysis is performed on the time series data of the resource consumption level to determine the optimization priority. Based on the optimization priority, a resource allocation adjustment plan for the blood glucose monitoring system is generated, and the change process of the resource consumption level of the blood glucose monitoring system is recorded to provide a reference for subsequent optimization decisions of the blood glucose monitoring system.

7. The method for real-time blood glucose monitoring with dynamic feedback correction according to claim 1, characterized in that, Step S5, which involves adjusting the blood glucose monitoring feedback interval to the target interval value in stable mode and determining it as the final feedback rhythm for the current monitoring cycle, includes: If it is determined that there is a need to optimize resource consumption in the blood glucose monitoring system, the blood glucose monitoring system is triggered to switch to a stable monitoring mode. Through an adaptive adjustment algorithm, the target feedback interval value in the stable monitoring mode is calculated. Based on the target feedback interval value, the collection cycle of interstitial fluid data and the processing cycle of blood glucose data are extended. By balancing the monitoring accuracy requirements and resource utilization efficiency, the adjustment range of the feedback interval is optimized. If the resource consumption level of the blood glucose monitoring system returns to the preset normal range after the feedback interval is adjusted, the current feedback interval is determined as the final feedback rhythm, and the relevant parameters of the final feedback rhythm are recorded to provide data basis for subsequent dynamic adjustment of the feedback interval.

8. The method for real-time blood glucose monitoring with dynamic feedback correction according to claim 1, characterized in that, The step S6, updating the monitoring parameters of the blood glucose monitoring system to adapt the system to individual physiological differences and dynamic changes in blood glucose, includes: Based on the final feedback rhythm, adjust the operating parameters of the blood glucose monitoring system to match the current blood glucose monitoring needs, collect and analyze individual physiological difference data of users, customize and update the operating parameters of the blood glucose monitoring system, and dynamically adjust the configuration strategy of the operating parameters of the blood glucose monitoring system according to the real-time trend of blood glucose dynamic changes. Verify the monitoring stability of the blood glucose monitoring system after adjusting the operating parameters, and determine the optimization effect of the operating parameters after adjustment based on the verification results. If the optimization effect meets the preset requirements, save the current blood glucose monitoring system operating parameters as the default configuration of the blood glucose monitoring system, record the adjustment history of the blood glucose monitoring system operating parameters, and provide reference data for subsequent personalized optimization.

9. A dynamic feedback correction real-time blood glucose monitoring system, used to implement the dynamic feedback correction real-time blood glucose monitoring method as described in any one of claims 1-8, characterized in that, The system includes: The deviation determination unit is used to acquire interstitial fluid data and venous blood reference values, and determine the current sensor drift deviation based on the real-time difference between the two. A rate calculation unit is used to correct the interstitial fluid data based on the drift deviation of the sensor, and to calculate the blood glucose fluctuation rate based on the corrected interstitial fluid data. The mode triggering unit is used to compare the blood glucose fluctuation rate with a preset fluctuation threshold. If the blood glucose fluctuation rate exceeds the preset fluctuation threshold, the blood glucose monitoring feedback interval is adjusted to a high-frequency mode. The result generation unit is used to perform calibration correction based on the adjusted feedback interval in the high-frequency mode, and to correct the interstitial fluid data by fusing the venous blood reference value, thereby generating a corrected blood glucose monitoring result. The resource adjustment unit is used to evaluate the resource consumption level of the blood glucose monitoring system based on the corrected blood glucose monitoring results. If there is a need to optimize resource consumption, the feedback interval of blood glucose monitoring is adjusted to the target interval value in the stable mode and determined as the final feedback rhythm of the current monitoring cycle. The parameter adaptation unit is used to update the monitoring parameters of the blood glucose monitoring system based on the final feedback rhythm, so that the blood glucose monitoring system can adapt to individual physiological differences and dynamic changes in blood glucose.

10. A computer-readable storage medium storing instructions thereon, characterized in that, When the instruction is executed by the processor, it implements a real-time blood glucose monitoring method with dynamic feedback correction as described in any one of claims 1-8.

Images