Blood glucose monitoring control system and control method

Abstract

The invention relates to the technical field of data analysis, and discloses a blood glucose monitoring control system and a control method, which realize accurate dynamic monitoring of a blood glucose state through acquisition, alignment and fusion of multi-source physiological signals. The system firstly collects blood glucose, skin temperature, acceleration and skin electric signals, and performs time compensation and filtering processing to ensure data synchronization. Inertia correction is calculated according to changes of blood glucose and skin temperature, delay and physiological inertia influences are eliminated, and abnormal jumps are recognized. The interference intensity is calculated in combination with movement and galvanic skin change, and the blood glucose value is corrected. The stability of the system is dynamically adjusted by analyzing the mean value and variance of recent blood glucose data. Short-term trend prediction is generated by combining the blood glucose change rate and the acceleration, and dynamic correction is performed when the change is severe. The individual risk is evaluated according to the predicted deviation from the current blood sugar, and early warning or intervention prompt is automatically sent out when the high-low blood sugar risk occurs.

Description

Technical Field

[0001] This invention relates to the field of data analysis technology, specifically to a blood glucose monitoring and control system and method. Background Technology

[0002] Diabetes mellitus is a chronic metabolic disease caused by abnormal insulin secretion or action. Patients need to monitor their blood glucose levels long-term to prevent complications from hyperglycemia or hypoglycemia. Traditional blood glucose monitoring methods mainly rely on finger-prick blood sampling. While this method is highly accurate, it has drawbacks such as cumbersome operation, poor real-time performance, and inability to continuously reflect blood glucose fluctuations, resulting in low patient compliance. In recent years, the application of continuous glucose monitoring devices has gradually become more widespread. These devices can achieve continuous sampling through subcutaneous sensors, but in practical use, problems such as signal delay, measurement noise, temperature drift, motion interference, and individual differences still exist, leading to insufficient dynamic accuracy and predictive precision of blood glucose data.

[0003] Existing technologies typically employ simple filtering or interpolation algorithms to smooth acquired signals, failing to effectively correct for asynchronous sampling from different physiological sensors and lacking adaptive compensation mechanisms for delay and inertial errors. Furthermore, external factors such as patient movement, changes in skin temperature, and fluctuations in skin conductance signals significantly affect the stability of blood glucose sensors, causing momentary jumps or abnormal drifts in monitoring results, making it difficult to accurately reflect the true trend of blood glucose changes. In addition, while some systems can retrospectively analyze historical data, they lack real-time short-term predictive capabilities and risk assessment mechanisms, failing to promptly identify the risk of impending hypoglycemia or hyperglycemia, thus delaying intervention.

[0004] In existing research, blood glucose monitoring algorithms mostly use fixed weights or empirical parameters for filtering and prediction, resulting in poor interpretability and a lack of universality. When a patient's physiological state, external temperature, or exercise intensity changes, these empirical parameters often cannot be adjusted in a timely manner, leading to unstable monitoring results. Furthermore, different sensors vary significantly in response speed and sampling accuracy. Without a unified time alignment and data fusion mechanism, information shifts and distortions between signals will occur, making dynamic monitoring results unreliable.

[0005] Therefore, there is an urgent need for a dynamic blood glucose monitoring system that can integrate multi-source physiological signals and possesses functions such as time synchronization, inertial decoupling, interference inversion, self-correction, and risk prediction. This system should, based on accurate alignment of multiple signals, automatically identify and correct abnormal signals through comprehensive analysis of blood glucose, skin temperature, skin conductance, and motion information, ensuring a smooth blood glucose curve that reflects the true physiological state. Simultaneously, the system should also possess self-awareness capabilities, enabling self-correction based on data stability to eliminate accumulated errors. Furthermore, it should proactively identify abnormal blood glucose trends through dynamic trend prediction and risk assessment modules, providing early warnings and intervention prompts, thereby improving the accuracy and safety of blood glucose management for diabetic patients. Summary of the Invention

[0006] This invention provides a blood glucose monitoring and control system and method, which helps to solve the problems mentioned in the background art.

[0007] This invention provides the following technical solution: a blood glucose monitoring and control system, comprising:

[0008] The multi-source physiological signal acquisition and time alignment module collects data from blood glucose, skin temperature, acceleration and skin conductance sensors, and performs time synchronization and filtering on different signals;

[0009] The dynamic inertial decoupling and instantaneous change detection module calculates inertial correction based on the instantaneous changes in blood glucose and skin temperature, eliminates the influence of sensing delay and physiological inertia, and identifies and marks abnormal jumps.

[0010] The physiological interference inversion correction module integrates motion, skin conductance, and temperature information to calculate the interference intensity and correct blood glucose output.

[0011] The self-cognitive stability correction module calculates the system stability index by statistically analyzing recent blood glucose data and adjusts the net blood glucose value accordingly.

[0012] The short-term trend prediction and dynamic correction module generates short-term predictions based on the rate and acceleration of blood glucose changes, and performs dynamic corrections under acceleration or deceleration conditions.

[0013] The risk assessment and intervention triggering module assesses individual risk based on the deviation between predicted and current blood glucose levels, and generates intervention notifications or observation prompts when there is a risk of hypoglycemia or hyperglycemia.

[0014] Optionally, the process of acquiring data from blood glucose, skin temperature, acceleration, and skin conductance sensors, and performing time synchronization and filtering on different signals, includes:

[0015] Blood glucose, skin temperature, acceleration, and skin conductance data of patients are collected using blood glucose monitors, skin temperature sensing patches, wearable devices, and skin conductance monitors.

[0016] Time compensation is applied to sampled data with delays, and a unified time base is used to calculate the observation period.

[0017] Linear interpolation is performed on all types of sampled signals on a unified time base, and median filtering is performed in conjunction with the data from the current and the previous two observation periods.

[0018] Error tolerance is implemented for missing data, and an alert is triggered when all three values ​​are missing.

[0019] Optionally, the step of calculating inertial correction based on instantaneous changes in blood glucose and skin temperature to eliminate sensing delay and physiological inertia effects, and to identify and mark abnormal jumps, includes:

[0020] The instantaneous derivatives of blood glucose and skin temperature are calculated using the central difference method under a unified time reference to obtain the rates of change of blood glucose and temperature;

[0021] The inertia ratio is calculated and the cumulative inertia effect is eliminated by numerical integration to obtain the corrected blood glucose value;

[0022] The difference between the corrected blood glucose value and the previous observation period is determined.

[0023] If the threshold is exceeded, an abnormal jump is marked, and the original blood glucose and correction data are saved for subsequent manual review.

[0024] Optionally, the process of integrating motion, skin conductance, and temperature information to calculate interference intensity and correct blood glucose output includes:

[0025] Calculate the first derivatives of patient acceleration, skin conductance, and skin temperature on a unified time reference to quantify the interference of exercise and physiological changes on blood glucose measurement;

[0026] The blood glucose level is corrected using the interference intensity value to obtain the net blood glucose value.

[0027] Optionally, the step of calculating the system stability index by statistically analyzing recent blood glucose data and adjusting the net blood glucose value includes:

[0028] In each observation period, take the most recent net blood glucose samples and calculate the mean and variance to generate a system stability index.

[0029] Determine whether correction is needed by analyzing the stability difference;

[0030] If necessary, calculate the correction increment and adjust the net blood glucose value. At the same time, limit the correction value that exceeds the limit, and set the conditions for continuous limit correction to trigger the automatic correction pause mechanism, requiring the patient to perform fingertip measurement and upload the results.

[0031] Optionally, the step of generating short-term predictions based on blood glucose change rate and acceleration, and dynamically correcting them under acceleration or deceleration conditions, includes:

[0032] Based on the first-order difference slope and second-order acceleration of net blood glucose over a continuous observation period, a short-term baseline prediction value of blood glucose is generated.

[0033] When a state of accelerated or decelerated blood glucose change is detected, a second-order acceleration correction is introduced into the baseline prediction to obtain the accelerated-corrected predicted blood glucose value.

[0034] Optionally, the step of assessing individual risk based on the deviation between predicted and current blood glucose levels, and generating intervention notifications or observation prompts when there is a risk of hypoglycemia or hyperglycemia, includes:

[0035] Based on the deviation between the accelerated correction of the predicted blood glucose and the current net blood glucose value, a relative risk index is calculated, and the patient's blood glucose status is graded and judged in conjunction with the blood glucose health index.

[0036] When predicting the risk of hypoglycemia or hyperglycemia, the system generates intervention notifications or observation prompts;

[0037] When the risk level is intermediate, it is recommended to continue monitoring.

[0038] The present invention has the following beneficial effects:

[0039] 1. By employing time compensation and interpolation methods, the system synchronizes and aligns data from different sensors, enabling dynamic blood glucose monitoring under the same time reference. Compared to traditional methods relying solely on a single blood glucose monitor, this approach considers both sampling delay and device asynchrony. Time correction is performed immediately after data acquisition, ensuring that each sampling point reflects the patient's true physiological state at that moment. The module also performs median filtering on multi-cycle data, effectively suppressing occasional noise and sensing errors, ensuring data stability. When a signal loss is detected, the system automatically triggers an alert and notifies staff, preventing data continuity interruptions from affecting subsequent calculations. Through a multi-channel, multi-time-series fusion mechanism, the system achieves high-precision, high-time-efficiency signal integration, laying the foundation for subsequent blood glucose trend identification and interference compensation, significantly improving the overall reliability and practicality of the monitoring system.

[0040] 2. Based on the instantaneous change rates of blood glucose levels and skin temperature, an inertial correction model reflecting the relationship between human metabolic response and sensor delay is constructed, thereby eliminating dynamic errors caused by factors such as sensor response lag and subcutaneous fluid metabolism lag at the data level. Through the combination of differential and integral methods, the system can automatically identify abnormal jumps, preventing short-term fluctuations from misleading monitoring results. When an abnormal change is detected exceeding a set threshold, the system automatically marks and retains both the original and corrected data for manual verification by doctors. Unlike existing technologies that rely on fixed-time smoothing or single-point correction, this invention achieves real-time inertial compensation in the time domain, making the dynamic response of the blood glucose change curve closer to the actual physiological state. This design significantly improves the accuracy of continuous blood glucose monitoring, especially in complex scenarios such as exercise, body temperature fluctuations, or slight sensor offsets, maintaining smooth and reliable data.

[0041] 3. By comprehensively analyzing changes in acceleration, skin conductance, and temperature, an interference intensity model for blood glucose measurement was established. This model can identify blood glucose signal deviations caused by external physiological interferences such as movement and changes in skin conductivity in real time, and dynamically correct the blood glucose output accordingly. By establishing a quantitative correspondence between signal change rates, the system can distinguish between true blood glucose changes and external disturbances, thus making the net blood glucose value more accurate and reliable. Compared with traditional methods that rely solely on signal smoothing or threshold filtering, this invention implements an active compensation mechanism based on physiological parameter inversion. This mechanism can maintain data stability even under patient movement or drastic changes in ambient temperature, significantly reducing false alarms and misjudgments. The introduction of this module enables the monitoring system to maintain high-precision output even in non-resting states, thereby ensuring the effectiveness and clinical usability of continuous monitoring.

[0042] 4. By statistically analyzing blood glucose data over a continuous observation period, the system calculates a stability index and automatically performs self-correction based on fluctuation characteristics. The system can determine in real time whether data stability has decreased and corrects the current blood glucose value by a limited margin when abnormal fluctuations are detected, thereby avoiding error accumulation. If continuous over-limit corrections occur, the system will pause automatic correction and prompt for manual intervention to ensure safety. Unlike traditional filtering algorithms based on static thresholds, this solution implements a self-aware blood glucose data adjustment mechanism through dynamic statistical modeling. It can adaptively adjust the output according to changes in the patient's condition, making the monitoring results closer to the true physiological trend. This module enables the system to maintain data consistency during long-term continuous operation, reduces the number of manual calibrations, and improves the sustainable use performance of the device in home and clinical settings.

[0043] 5. By calculating the rate and acceleration of blood glucose changes, high-precision prediction of short-term blood glucose trends is achieved. When blood glucose changes are detected to be accelerating or decelerating, the system automatically introduces a second-order correction term to make the predicted value more consistent with the dynamic characteristics of human metabolism. This module can identify rapid blood glucose change trends before obvious abnormalities appear, providing medical staff and patients with an opportunity for early intervention. Compared with the traditional linear extrapolation method, this solution uses dynamic differential relationships to establish an adaptive trend prediction model, which not only improves prediction accuracy but also significantly enhances the system's early warning response capability. In practical applications, this module can effectively reduce the delayed response to sudden hypoglycemia or hyperglycemia events, improving the safety protection level of the monitoring system in clinical and home scenarios.

[0044] 6. By comprehensively analyzing the difference between the predicted and current blood glucose levels, individualized risk indicators are calculated, and the patient's blood glucose status is classified and assessed using a health index function. When the system detects that the probability of low or high blood glucose exceeds the safety threshold, it immediately generates an intervention instruction to notify medical staff or the patient to perform the corresponding operation. If the risk is in the intermediate range, the system will prompt continued observation, thereby avoiding over-intervention. This module realizes a closed-loop management mechanism of risk grading and intelligent response. Unlike traditional passive monitoring methods, this invention can proactively identify potential risks and make dynamic decisions, enabling blood glucose management to shift from single monitoring to intelligent prediction and proactive intervention. This design significantly improves patient safety assurance capabilities, reduces the clinical risks caused by delayed responses, and provides highly reliable support for long-term intelligent management of diabetes. Attached Figure Description

[0045] Figure 1 This is a schematic diagram of the process of the present invention.

[0046] Figure 2 This is a schematic diagram of the system structure of the present invention. Detailed Implementation

[0047] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0048] Example 1, see Figures 1 to 2 A blood glucose monitoring and control system, comprising:

[0049] The multi-source physiological signal acquisition and time alignment module collects data from blood glucose, skin temperature, acceleration and skin conductance sensors, and performs time synchronization and filtering on different signals;

[0050] The dynamic inertial decoupling and instantaneous change detection module calculates inertial correction based on the instantaneous changes in blood glucose and skin temperature, eliminates the influence of sensing delay and physiological inertia, and identifies and marks abnormal jumps.

[0051] The physiological interference inversion correction module integrates motion, skin conductance, and temperature information to calculate the interference intensity and correct blood glucose output.

[0052] The self-cognitive stability correction module calculates the system stability index by statistically analyzing recent blood glucose data and adjusts the net blood glucose value accordingly.

[0053] The short-term trend prediction and dynamic correction module generates short-term predictions based on the rate and acceleration of blood glucose changes, and performs dynamic corrections under acceleration or deceleration conditions.

[0054] The risk assessment and intervention triggering module assesses individual risk based on the deviation between predicted and current blood glucose levels, and generates intervention notifications or observation prompts when there is a risk of hypoglycemia or hyperglycemia.

[0055] The process of acquiring data from blood glucose, skin temperature, acceleration, and skin conductance sensors, and performing time synchronization and filtering on different signals, includes:

[0056] Blood glucose, skin temperature, acceleration, and skin conductance data of patients are collected using blood glucose monitors, skin temperature sensing patches, wearable devices, and skin conductance monitors.

[0057] Time compensation is applied to sampled data with delays, and a unified time base is used to calculate the observation period.

[0058] Linear interpolation is performed on all types of sampled signals on a unified time base, and median filtering is performed in conjunction with the data from the current and the previous two observation periods.

[0059] Error tolerance is implemented for missing data, and an alert is triggered when all three values ​​are missing.

[0060] The calculation of inertial correction based on instantaneous changes in blood glucose and skin temperature eliminates the effects of sensing delay and physiological inertia, and identifies and marks abnormal jumps, including:

[0061] The instantaneous derivatives of blood glucose and skin temperature are calculated using the central difference method under a unified time reference to obtain the rates of change of blood glucose and temperature;

[0062] The inertia ratio is calculated and the cumulative inertia effect is eliminated by numerical integration to obtain the corrected blood glucose value;

[0063] The difference between the corrected blood glucose value and the previous observation period is determined.

[0064] If the threshold is exceeded, an abnormal jump is marked, and the original blood glucose and correction data are saved for subsequent manual review.

[0065] The integrated motion, skin conductance, and temperature information are used to calculate interference intensity and correct blood glucose output, including:

[0066] Calculate the first derivatives of patient acceleration, skin conductance, and skin temperature on a unified time reference to quantify the interference of exercise and physiological changes on blood glucose measurement;

[0067] The blood glucose level is corrected using the interference intensity value to obtain the net blood glucose value.

[0068] The process of calculating the system stability index by statistically analyzing recent blood glucose data and adjusting the net blood glucose value includes:

[0069] In each observation period, take the most recent net blood glucose samples and calculate the mean and variance to generate a system stability index.

[0070] Determine whether correction is needed by analyzing the stability difference;

[0071] If necessary, calculate the correction increment and adjust the net blood glucose value. At the same time, limit the correction value that exceeds the limit, and set the conditions for continuous limit correction to trigger the automatic correction pause mechanism, requiring the patient to perform fingertip measurement and upload the results.

[0072] The method of generating short-term predictions based on blood glucose change rate and acceleration, and dynamically correcting them under acceleration or deceleration conditions, includes:

[0073] Based on the first-order difference slope and second-order acceleration of net blood glucose over a continuous observation period, a short-term baseline prediction value of blood glucose is generated.

[0074] When a state of accelerated or decelerated blood glucose change is detected, a second-order acceleration correction is introduced into the baseline prediction to obtain the accelerated-corrected predicted blood glucose value.

[0075] The method of assessing individual risk based on the deviation between predicted and current blood glucose levels, and generating intervention notifications or observation prompts when there is a risk of hypoglycemia or hyperglycemia, includes:

[0076] Based on the deviation between the accelerated correction of the predicted blood glucose and the current net blood glucose value, a relative risk index is calculated, and the patient's blood glucose status is graded and judged in conjunction with the blood glucose health index.

[0077] When predicting the risk of hypoglycemia or hyperglycemia, the system generates intervention notifications or observation prompts;

[0078] When the risk level is intermediate, it is recommended to continue monitoring.

[0079] Example 2: A blood glucose monitoring and control system, comprising:

[0080] The multi-source physiological signal acquisition and time alignment module collects data from blood glucose, skin temperature, acceleration and skin conductance sensors, and performs time synchronization and filtering on different signals;

[0081] The dynamic inertial decoupling and instantaneous change detection module calculates inertial correction based on the instantaneous changes in blood glucose and skin temperature, eliminates the influence of sensing delay and physiological inertia, and identifies and marks abnormal jumps.

[0082] The physiological interference inversion correction module integrates motion, skin conductance, and temperature information to calculate the interference intensity and correct blood glucose output.

[0083] The self-cognitive stability correction module calculates the system stability index by statistically analyzing recent blood glucose data and adjusts the net blood glucose value accordingly.

[0084] The short-term trend prediction and dynamic correction module generates short-term predictions based on the rate and acceleration of blood glucose changes, and performs dynamic corrections under acceleration or deceleration conditions.

[0085] The risk assessment and intervention triggering module assesses individual risk based on the deviation between predicted and current blood glucose levels, and generates intervention notifications or observation prompts when there is a risk of hypoglycemia or hyperglycemia.

[0086] The process of acquiring data from blood glucose, skin temperature, acceleration, and skin conductance sensors, and performing time synchronization and filtering on different signals, includes:

[0087] Read the sampling time of the blood glucose monitor ;

[0088] The sampling time for obtaining the patient's skin surface temperature was obtained using a skin temperature sensing patch. ;

[0089] The sampling time of patient acceleration was collected using wearable devices. ;

[0090] The skin conductance / resistance sampling time of the patient was obtained using a skin conductance monitor. ;

[0091] If there is a delay when the sampling device returns. Then, time compensation is performed immediately during reading: the original time is replaced with... ;

[0092] Indicates data type;

[0093] The sampling equipment includes a blood glucose monitor, a skin temperature sensing patch, a wearable device, and a skin conductivity monitor;

[0094] Set the observation period and calculate the unified time base for the current observation period. :

[0095] ;

[0096] The sampling time for each type of sampling device is within Perform linear interpolation at:

[0097] ;

[0098] in:

[0099] This represents the nth data collection time.

[0100] Represents the original sampled value;

[0101] This indicates that a unified time base is obtained after linear interpolation. The alignment value;

[0102] Will Median filtering is applied to the current and the two previous observation periods, outputting a smoothed value. :

[0103] ;

[0104] Where median represents the median. , These represent the unified time references for the first and second observation periods prior to the current observation period, respectively.

[0105] like If any value is missing, the median is taken from the remaining non-missing values;

[0106] like If all three values ​​are missing, an alert will be triggered immediately and staff will be notified.

[0107] Synchronization and alignment of data from different sensors are achieved through time compensation and interpolation methods, enabling dynamic blood glucose monitoring under the same time reference. Compared to traditional methods relying solely on a single blood glucose monitor, this method considers both sampling delay and device asynchrony. By performing time correction immediately after data acquisition, each sampling point reflects the patient's true physiological state at that moment. The module also performs median filtering on multi-period data, effectively suppressing occasional noise and sensing errors, ensuring data stability. When a signal loss is detected, the system automatically triggers an alert and notifies staff, thus preventing data continuity interruptions from affecting subsequent calculations. Through a multi-channel, multi-time-series fusion mechanism, the system achieves high-precision, high-time-efficiency signal integration, laying the foundation for subsequent blood glucose trend identification and interference compensation, significantly improving the overall reliability and practicality of the monitoring system.

[0108] The calculation of inertial correction based on instantaneous changes in blood glucose and skin temperature eliminates the effects of sensing delay and physiological inertia, and identifies and marks abnormal jumps, including:

[0109] Calculate the instantaneous derivatives of the patient's blood glucose and skin surface temperature using the central difference method:

[0110] Find the difference between blood glucose levels:

[0111] ;

[0112] in, This indicates the patient's blood glucose measurement value. Indicates the time step. Indicates in The instantaneous derivative of blood glucose at the point;

[0113] Calculate the difference in skin surface temperature:

[0114] ;

[0115] in, This indicates the patient's skin surface temperature value. Indicates in Estimates of the instantaneous derivative of skin temperature at that location;

[0116] Determine whether the instantaneous derivative estimate of the patient's skin surface temperature is usable:

[0117] like If so, then set the inertia ratio to zero;

[0118] in, The threshold for numerical stability;

[0119] The above inertia ratio Specifically:

[0120] ;

[0121] Eliminating the inertial accumulation term using numerical integration :

[0122] ;

[0123] Instantaneous blood glucose after inertia correction ;

[0124] This refers to the initial time when the sampling device collects data.

[0125] right Perform single-step anomaly detection:

[0126] If the current observation period The difference from the previous observation period is greater than the set threshold. If so, mark the current situation as an abnormal transition and save the original data. and For subsequent manual review.

[0127] Based on the instantaneous change rates of blood glucose levels and skin temperature, an inertial correction model reflecting the relationship between human metabolic response and sensor delay is constructed, thereby eliminating dynamic errors caused by factors such as sensor response lag and subcutaneous fluid metabolism lag at the data level. Through a combination of differential and integral methods, the system can automatically identify abnormal fluctuations, preventing short-term fluctuations from misleading monitoring results. When an abnormal change exceeds a set threshold, the system automatically marks and retains both the original and corrected data for manual verification by doctors. Unlike existing technologies that rely on fixed-time smoothing or single-point correction, this invention achieves real-time inertial compensation in the time domain, making the dynamic response of the blood glucose change curve closer to the actual physiological state. This design significantly improves the accuracy of continuous blood glucose monitoring, especially in complex scenarios such as exercise, body temperature fluctuations, or slight sensor offsets, maintaining smooth and reliable data.

[0128] The integrated motion, skin conductance, and temperature information are used to calculate interference intensity and correct blood glucose output, including:

[0129] Using central difference calculation acceleration Skin resistance Skin surface temperature The first derivative of is specifically expressed as ;

[0130] Calculate interference strength :

[0131] ;

[0132] in:

[0133] Let it be a constant, and take the AND... The smallest positive number with the same dimension is used to prevent numerical instability;

[0134] Calculate net blood glucose measurement value .

[0135] By comprehensively analyzing changes in acceleration, skin conductance, and temperature, an interference intensity model for blood glucose measurement was established. This model can identify blood glucose signal deviations caused by external physiological disturbances such as movement and changes in skin conductivity in real time, and dynamically correct the blood glucose output accordingly. By establishing a quantitative correspondence between signal change rates, the system can distinguish between true blood glucose changes and external disturbances, thus making the net blood glucose value more accurate and reliable. Compared with traditional methods that rely solely on signal smoothing or threshold filtering, this invention implements an active compensation mechanism based on physiological parameter inversion. This mechanism can maintain data stability even under patient movement or drastic changes in ambient temperature, significantly reducing false alarms and misjudgments. The introduction of this module enables the monitoring system to maintain high-precision output even in non-resting states, thereby ensuring the effectiveness and clinical usability of continuous monitoring.

[0136] The process of calculating the system stability index by statistically analyzing recent blood glucose data and adjusting the net blood glucose value includes:

[0137] exist Take the nearest Net blood glucose sample And calculate its mean. With variance :

[0138] ;

[0139] Where Mean represents the average value operation and Var represents the variance operation;

[0140] According to the mean With variance Calculate the stability index ;

[0141] in, It is a constant, and its value is 0.1 to avoid the denominator being zero;

[0142] Calculate the stability difference ;

[0143] in, The sampling point interval;

[0144] like Therefore, it is determined that self-correction is not necessary at present.

[0145] Otherwise, calculate the correction increment using the following formula. And perform the correction:

[0146] ;

[0147] in, It is a constant, and its value is 0.01 to avoid the denominator being zero;

[0148] Output corrected net blood glucose value ;

[0149] like Exceeding the absolute threshold Then Marked as limit correction;

[0150] If three consecutive calibrations are limited corrections, the automatic calibration function will be paused and the patient will be required to perform a fingertip measurement using a lancet and upload the result. Automatic calibration will only be restored after the doctor reviews the result.

[0151] By statistically analyzing blood glucose data over a continuous observation period, the system calculates a stability index and automatically performs self-correction based on fluctuation characteristics. The system can determine in real time whether data stability has decreased and corrects the current blood glucose value by a limited margin when abnormal fluctuations are detected, thus avoiding error accumulation. If continuous over-limit corrections occur, the system will pause automatic correction and prompt for manual intervention to ensure safety. Unlike traditional static threshold-based filtering algorithms, this solution implements a self-aware blood glucose data adjustment mechanism through dynamic statistical modeling. It can adaptively adjust the output according to changes in the patient's condition, making the monitoring results closer to the actual physiological trend. This module enables the system to maintain data consistency during long-term continuous operation, reduces the number of manual calibrations, and improves the sustainable use performance of the device in home and clinical settings.

[0152] The method of generating short-term predictions based on blood glucose change rate and acceleration, and dynamically correcting them under acceleration or deceleration conditions, includes:

[0153] Calculate the first-order difference slope of the current net blood glucose. :

[0154] ;

[0155] Calculate the second-order acceleration term :

[0156] ;

[0157] Indicates the acceleration threshold;

[0158] Based on the first difference slope of net blood glucose Generate basic predictions :

[0159] ;

[0160] in, Indicates the prediction time interval;

[0161] like Then mark it as "accelerated state" and add a second-order correction to the basic prediction:

[0162] ;

[0163] in, This indicates the predicted blood glucose value after acceleration correction.

[0164] By calculating the rate and acceleration of blood glucose changes, this module achieves high-precision prediction of short-term blood glucose trends. When blood glucose changes are detected to be accelerating or decelerating, the system automatically introduces a second-order correction term to make the predicted value more consistent with the dynamic characteristics of human metabolism. This module can identify rapid blood glucose change trends before obvious abnormalities appear, providing medical staff and patients with an opportunity for early intervention. Compared to traditional linear extrapolation methods, this solution utilizes dynamic differential relationships to establish an adaptive trend prediction model, which not only improves prediction accuracy but also significantly enhances the system's early warning response capability. In practical applications, this module can effectively reduce delayed responses to sudden hypoglycemia or hyperglycemia events, improving the safety and protection level of the monitoring system in clinical and home settings.

[0165] The method of assessing individual risk based on the deviation between predicted and current blood glucose levels, and generating intervention notifications or observation prompts when there is a risk of hypoglycemia or hyperglycemia, includes:

[0166] Calculate the patient's relative risk index :

[0167] ;

[0168] in, It is a constant, and its value is 1 to prevent the denominator from being zero;

[0169] Calculate the patient's blood glucose health index :

[0170] ;

[0171] in, This represents a mathematical function used to calculate the cumulative probability of a given standardized variable under a standard normal distribution.

[0172] like If the blood glucose level is low, it will immediately generate a notification for staff to intervene.

[0173] like and If so, the patient is advised to continue observation;

[0174] like and If this is detected, it indicates that the patient's blood sugar level is too high, and a notification will be immediately generated to inform staff for intervention.

[0175] By comprehensively analyzing the difference between the predicted and current blood glucose levels, individualized risk indicators are calculated, and the patient's blood glucose status is classified and assessed using a health index function. When the system detects that the probability of low or high blood glucose exceeds a safety threshold, it immediately generates an intervention instruction to notify medical staff or the patient to perform the corresponding action. If the risk is in the intermediate range, the system will prompt continued observation, thereby avoiding over-intervention. This module realizes a closed-loop management mechanism of risk grading and intelligent response. Unlike traditional passive monitoring methods, this invention can proactively identify potential risks and make dynamic decisions, shifting blood glucose management from simple monitoring to intelligent prediction and proactive intervention. This design significantly improves patient safety, reduces the clinical risks caused by delayed responses, and provides highly reliable support for long-term intelligent management of diabetes.

[0176] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0177] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A blood glucose monitoring control system, characterized by, The method comprises the following steps: A multi-source physiological signal acquisition and time alignment module collects data from blood glucose, skin temperature, acceleration and skin conductance sensors, and performs time synchronization and filtering processing on different signals; A dynamic inertia decoupling and transient change detection module calculates inertia correction according to the transient changes of blood glucose and skin temperature, eliminates the influence of sensor delay and physiological inertia, and identifies and marks abnormal jumps; A physiological interference inversion correction module calculates the interference intensity and corrects the blood glucose output by integrating motion, skin conductance and temperature information; A self-cognition stability correction module calculates the system stability index by statistical recent blood glucose data, and adjusts the net blood glucose value; A short-term trend prediction and dynamic correction module generates a short-term prediction based on the blood glucose change rate and acceleration, and performs dynamic correction in the acceleration or deceleration state; A risk assessment and intervention triggering module assesses the individual risk according to the deviation of predicted blood glucose and current blood glucose, and generates intervention notification or observation prompt when there is a risk of hypoglycemia or hyperglycemia.

2. The control method of a blood sugar monitoring control system according to claim 1, wherein The method comprises the following steps: Collecting blood glucose, skin temperature, acceleration and skin conductance data of the patient through a blood glucose monitor, a skin temperature sensing patch, a wearable device and a skin conductance monitor; Time compensation is performed on the sampling data with delay, and the time reference of the observation period is uniformly calculated; Linear interpolation processing is performed on various sampling signals on the unified time reference, and median filtering processing is performed in combination with the data of the current and previous two observation periods; Fault tolerance processing is performed on the missing data, and a warning is triggered when three values are missing.

3. The control method of a blood glucose monitoring control system according to claim 1, wherein The method comprises the following steps: The instantaneous derivative of blood glucose and skin temperature is calculated using central difference under the unified time reference to obtain the blood glucose and temperature change rate; The inertia ratio is calculated and the inertia accumulation effect is eliminated by numerical integration to obtain the corrected blood glucose value; Difference judgment is performed on the blood glucose correction value and the previous observation period; If the set threshold is exceeded, an abnormal jump is marked, and the original blood glucose and correction data are saved for subsequent manual review processing.

4. The control method of a blood sugar monitoring control system according to claim 1, wherein The method comprises the following steps: The first derivative of the patient's acceleration, skin conductance and skin temperature under the unified time reference is calculated to quantify the interference of motion and physiological changes on blood glucose measurement; The blood glucose is corrected using the interference intensity value to obtain the net blood glucose value.

5. The control method of a blood sugar monitoring control system according to claim 1, wherein, The method comprises the following steps: The mean and variance are calculated to generate the system stability index by taking the latest several net blood glucose samples in each observation period; Whether correction is needed is judged by the stability difference; If needed, the correction increment is calculated and the net blood glucose value is adjusted, while the correction value exceeding the limit is limited, and the automatic correction suspension mechanism is triggered under the condition of continuous limited correction, requiring the patient to perform fingertip measurement and upload the results.

6. The control method of a blood sugar monitoring control system according to claim 1, wherein, The method comprises the following steps: Based on the first-order difference slope and the second-order acceleration of net blood glucose in the continuous observation period, a short-term blood glucose base prediction value is generated; In the state of detecting blood glucose acceleration or deceleration, the second-order acceleration correction is introduced into the base prediction to obtain the prediction blood glucose value after acceleration correction.

7. The control method of a blood sugar monitoring control system according to claim 1, wherein The individual risk is assessed according to the deviation of the predicted blood glucose from the current blood glucose, and intervention notification or observation prompt is generated when there is a risk of hypoglycemia or hyperglycemia, including: According to the deviation of the predicted blood glucose after acceleration correction from the current net blood glucose value, the relative risk index is calculated, and the blood glucose state of the patient is judged by grading combined with the blood glucose health index; When predicting the risk of hypoglycemia or hyperglycemia, the system generates intervention notification or observation prompt; In the intermediate risk state, continue to observe.

Images