Chronic disease health management method based on key index correlation analysis

Abstract

The invention relates to the technical field of brain chronic disease health management, in particular to a chronic disease health management method based on key index correlation analysis, which comprises the following steps: acquiring a historical review data set of a patient, and synchronously acquiring continuous blood glucose monitoring data and micro cognitive game interaction data within a preset review time period; and a complete time sequence record is formed by the real-time blood glucose value with the timestamp, the execution function score and the cognitive task load coefficient. A correlation model capable of capturing complex interaction between blood glucose and cognitive effectiveness is constructed by performing time sequence alignment and multi-dimensional mapping on a historical review data set. The core of the correlation model is to reversely derive an individualized functional blood glucose threshold interval by using peak distribution of execution function scores. According to the method, the limitation of a traditional fixed blood glucose standard is broken through, blood glucose regulation and control with the optimal cognitive expression of the patient as the benchmark are achieved, and a new scientific basis is provided for personalized health management.

Description

Technical Field

[0001] This invention belongs to the field of chronic brain disease health management technology, and relates to a chronic disease health management method based on key indicator correlation analysis. Background Technology

[0002] Cerebrovascular diseases are among the leading causes of disability and death worldwide. Their occurrence and development are often accompanied by impaired blood flow to the brain, which in turn affects various higher neurological functions such as cognition, movement, and sensation, posing a serious threat to patients' quality of life. Meanwhile, blood glucose, as an important energy source for the body, plays an irreplaceable role in maintaining a stable level for the normal metabolism and function of all organs and tissues. This is especially true for the brain, which has extremely high energy demands; the dynamic balance of blood glucose levels directly affects the survival and activity efficiency of nerve cells.

[0003] It is worth noting that the brain, as an organ highly dependent on glucose for energy, is far more sensitive to blood glucose fluctuations than other tissues. Long-term or abrupt blood glucose abnormalities not only directly damage vascular endothelium and exacerbate arteriosclerosis, but may also indirectly affect the structure and function of cerebrovascular systems through mechanisms such as oxidative stress and inflammatory responses. This makes blood glucose management particularly significant in the prevention and control of cerebrovascular diseases. Especially for cerebrovascular patients with pre-existing glucose metabolism abnormalities, there is a complex and close interaction between blood glucose levels and their neurological function, cognitive performance, and disease prognosis.

[0004] However, current management of cerebrovascular disease patients with combined glucose metabolism abnormalities is significantly inadequate. Firstly, traditional monitoring systems rely excessively on standardized blood glucose ranges, neglecting individual differences. Clinical practice shows that even diabetic patients who take hypoglycemic drugs regularly and have well-controlled fasting blood glucose may still experience stroke due to the "chronic erosion" process of blood vessels. This reflects a major flaw in assessments that focus solely on blood glucose levels while ignoring vascular health. Blood glucose levels only assess whether a patient's blood glucose is high or low, failing to reflect the dynamic impact of blood glucose changes on cognitive performance and lacking a systematic analysis of cognitive impairment caused by abnormal blood glucose fluctuations. This results in existing health warning and intervention strategies being largely reactive, unable to provide targeted energy compensation recommendations before cognitive decline occurs, leading to a lag in overall chronic disease health management. Summary of the Invention

[0005] In view of the problems existing in the prior art, the present invention provides a chronic disease health management method based on key indicator correlation analysis to solve the above-mentioned technical problems.

[0006] To achieve the above and other objectives, the technical solution adopted by the present invention is as follows:

[0007] The first aspect of this invention provides a method for chronic disease health management based on key indicator correlation analysis. This method is applied to a data processing system deployed on a hospital server and is specifically designed for patients with cerebrovascular disease and diabetes. The method includes the following steps:

[0008] Obtain the patient's historical follow-up dataset, which includes: continuous blood glucose monitoring data and mini cognitive game interaction data collected synchronously within the preset follow-up time period. The continuous blood glucose monitoring data includes real-time blood glucose values ​​with time points, and the mini cognitive game interaction data includes performance function scores with time points and corresponding cognitive task load coefficients.

[0009] We performed time-series alignment and multidimensional mapping on the historical review dataset to construct a blood glucose cognitive efficacy association model. Based on the peak distribution of executive function scores in the model, we back-defined the functional blood glucose threshold range for patients.

[0010] Regression analysis was performed on continuous blood glucose monitoring data and cognitive task load coefficients to calculate the cognitive glucose consumption decay rate under different cognitive loads.

[0011] Fluctuation lag analysis was performed on historical review datasets to determine the impairment lag time parameter that caused the decline in executive function scores due to abnormal fluctuations in blood glucose levels.

[0012] Real-time monitoring of patients' current blood glucose levels and the real-time load coefficient of ongoing cognitive tasks; using cognitive glucose decay rate to predict future blood glucose change trajectories.

[0013] If the predicted blood glucose trajectory falls outside the functional blood glucose threshold range before the damage lag time parameter is reached, a health intervention warning signal containing energy compensation recommendations is generated.

[0014] A second aspect of the present invention provides a chronic disease health management device based on key indicator correlation analysis, including a processor, a memory, and a communication bus;

[0015] The memory stores a computer-readable program that can be executed by the processor;

[0016] The communication bus enables communication between the processor and the memory;

[0017] When the processor executes the computer-readable program, it performs steps in the chronic disease health management method based on key indicator correlation analysis as described in any one of the present invention.

[0018] As described above, the chronic disease health management method based on key indicator correlation analysis provided by this invention has at least the following beneficial effects:

[0019] 1. The chronic disease health management method based on key indicator correlation analysis provided by this invention acquires the patient's historical follow-up dataset and simultaneously collects continuous blood glucose monitoring data and mini-cognitive game interaction data within a preset follow-up time period. A complete time-series record is formed by timestamped real-time blood glucose values, executive function scores, and cognitive task load coefficients. By performing time-series alignment and multidimensional mapping on the historical follow-up dataset, a correlation model capable of capturing the complex interaction between blood glucose and cognitive efficacy is constructed. The core of this correlation model lies in using the peak distribution of executive function scores to inversely derive individualized functional blood glucose threshold ranges. This method breaks the limitations of traditional fixed blood glucose standards, achieving blood glucose regulation based on the patient's optimal cognitive performance, providing a new scientific basis for personalized health management. This not only improves the accuracy of detection but also provides an innovative means for monitoring cognitive function in patients with diabetes and cerebrovascular disease.

[0020] 2. This invention utilizes regression analysis of continuous blood glucose monitoring data and cognitive task load coefficients to calculate the cognitive glucose decay rate under various cognitive load conditions, and performs fluctuation lag analysis to determine the lag time parameter of the impairment of executive function scores caused by abnormal blood glucose fluctuations. By monitoring real-time blood glucose values ​​and cognitive task load coefficients, the system can predict blood glucose change trajectories and generate a health intervention warning signal containing energy compensation suggestions when a potential risk of hypoglycemia occurs. This provides a novel solution for timely intervention and ensuring the stability of cognitive efficacy, making brain function protection more effective and innovative in clinical applications. Attached Figure Description

[0021] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a schematic diagram of the logical connection of the method of the present invention.

[0023] Figure 2 This is a schematic diagram illustrating the construction logic of the blood glucose cognitive efficacy association model of the present invention.

[0024] Figure 3 This is a schematic diagram of the device structure connection of the present invention. Detailed Implementation

[0025] The following description, in conjunction with the implementation of this invention, is merely an example and illustration of the concept of this invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the inventive concept or exceed the scope defined in these claims, all of which should fall within the protection scope of this invention.

[0026] Overview of this application:

[0027] In current technologies, blood glucose management for diabetic patients largely relies on fixed target ranges, making it difficult to address individualized cognitive performance needs. Traditional methods, when explaining inter-individual differences in blood glucose fluctuations, neglect the dynamic impact of cognitive activity on blood glucose levels, making it difficult to accurately maintain an individual's optimal cognitive function. Furthermore, existing blood glucose monitoring systems typically lack energy compensation mechanisms for high-intensity cognitive activity, failing to anticipate rapid drops in blood glucose due to cognitive load, thus affecting the real-time nature and effectiveness of precise interventions.

[0028] To address the aforementioned issues, the study discovered a close interaction between blood glucose levels and cognitive performance, and established a "blood glucose-cognition" coupling model to achieve error compensation. The research revealed that the optimal blood glucose range for maintaining cognitive performance is closely related to an individual's physiological and activity levels, leading to the proposal of dynamically adjusting blood glucose detection thresholds based on cognitive load levels. To effectively predict and compensate for blood glucose consumption caused by cognitive activity, further experimental verification was conducted, incorporating the mapping relationship between cognitive task load and blood glucose fluctuations into the model parameter adjustment mechanism to form a closed-loop feedback system.

[0029] Specifically, this invention first simultaneously collects the patient's CGM data and scores from mini-cognitive games performed at different blood glucose levels. By analyzing the task load intensity of the cognitive games, the system calculates their potential impact on blood glucose levels. Simultaneously, by combining the blood glucose level change trajectory in the CGM data, a sensitivity curve characterizing the blood glucose-cognitive coupling relationship is constructed, identifying the functional blood glucose threshold for maintaining optimal cognitive performance. When the cognitive task load exceeds the set threshold, the system automatically switches to an energy decay prediction mode. This mode predicts the rate of blood glucose consumption under high-load cognitive activity and issues a glucose replenishment warning before a significant drop. During continuous monitoring, the system adjusts model parameters in real time based on the temporal changes in blood glucose levels through a feedback mechanism, forming a dynamic optimization closed loop. For cases where the load threshold is not exceeded, the standard inversion model continues to output the patient's blood glucose status.

[0030] Compared to existing technologies, traditional methods rely on fixed blood glucose standards and lack cognitive load compensation mechanisms, making it difficult to dynamically adapt to individual cognitive needs and activity changes. This innovative approach integrates blood glucose monitoring with cognitive load analysis, establishing a "blood glucose-cognition" coupled model to achieve dynamic adjustments in blood glucose management. Unlike existing static monitoring models, this approach can intelligently switch detection strategies based on real-time cognitive load levels and dynamically adjust model parameters through closed-loop feedback, significantly improving detection reliability and intervention timeliness in complex situations.

[0031] Through the above technical solution, this invention effectively overcomes the problem of neglecting the impact of cognitive load on blood glucose, improving the accuracy of individualized blood glucose management while ensuring real-time monitoring. The dynamic calibration mechanism for blood glucose thresholds combines the sensitivity of the blood glucose-cognitive mapping model with the long-term stability of adaptive adjustment functions, providing a reliable technical means for intelligent blood glucose management in diabetic patients, and is particularly suitable for precise health intervention scenarios under conditions of drastic changes in cognitive load.

[0032] After introducing the basic concept of the present invention, the embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0033] Example 1:

[0034] Please see Figure 1 As shown, a chronic disease health management method based on key indicator correlation analysis is applied to a data processing system deployed on a hospital server. For patients with cerebrovascular disease and diabetes, the method includes the following steps:

[0035] Obtain the patient's historical follow-up dataset, which includes: continuous blood glucose monitoring data and mini cognitive game interaction data collected synchronously within the preset follow-up time period. The continuous blood glucose monitoring data includes real-time blood glucose values ​​with time points, and the mini cognitive game interaction data includes performance function scores with time points and corresponding cognitive task load coefficients.

[0036] In the above method, the cognitive task load coefficient in the micro-cognitive game interaction data is determined in the following way:

[0037] Acquire the game difficulty level, game operation frequency, and game screen refresh rate of the mini cognitive game;

[0038] The game difficulty level, game operation frequency, and game screen refresh rate are normalized.

[0039] The numerical value representing the brain's computational intensity is obtained by weighted summation and used as the cognitive task load coefficient.

[0040] Specifically, the system first records the core parameters of the cognitive game in real time: game difficulty level, game operation frequency, and game screen refresh rate. Among them, the game difficulty level represents the complexity and requirements of the task, the game operation frequency represents the intensity of interaction per unit time, and the game screen refresh rate indicates the speed of visual updates and its degree of stimulation to cognition. These elements are transformed into standardized values ​​through normalization processing, so that data from different dimensions can be effectively compared and integrated.

[0041] Let the parameters be D representing the standardized value of game difficulty level, F representing the standardized value of game operation frequency, and R representing the standardized value of game screen refresh rate. The formula for calculating the cognitive task load coefficient C is:

[0042] ;

[0043] in, , , The weights for game difficulty level, operation frequency, and screen refresh rate are respectively used, and these weights are calibrated experimentally to reflect the relative contribution of each parameter to cognitive load. , , , , , These represent the historical minimum and maximum values ​​for the corresponding parameters, ensuring that the proportions of each parameter appropriately reflect its range of variation during normalization. The cognitive task load coefficient serves as a baseline input for dynamically adjusting the patient's blood glucose threshold, enabling the system to adjust monitoring and intervention strategies based on real-time cognitive load.

[0044] We performed time-series alignment and multidimensional mapping on historical review datasets to construct a blood glucose cognitive efficacy association model. Based on the peak distribution of executive function scores in the model, we inversely defined the functional blood glucose threshold range for patients.

[0045] In the above methods, such as Figure 2 As shown, a time-series alignment and multidimensional mapping were performed on the historical review dataset to construct a blood glucose cognitive efficacy association model, including:

[0046] Based on the interactive data of the mini cognitive game, reaction score and memory score were extracted according to different cognitive dimensions.

[0047] Reaction score and memory score were correlated with blood glucose values ​​at the same time to construct reaction-blood glucose sub-models and memory-blood glucose sub-models, respectively.

[0048] Obtain imaging data of cerebrovascular lesions in patients to determine the functional brain regions where the lesions are located;

[0049] Based on the regulatory weights of brain functional areas on reaction time and memory, the reaction time-glucose sub-model and the memory-glucose sub-model are weighted and fused to generate a glucose cognitive efficacy correlation model.

[0050] Specifically, the historical review dataset is first time-series aligned and multidimensionally mapped to construct a blood glucose cognitive efficacy correlation model. This process begins by parsing and normalizing the micro-cognitive game interaction data for different cognitive dimensions, mapping the original reaction test data and memory test data with time points to a unified [0,1] efficacy interval. Then, based on the sampling time points of continuous blood glucose monitoring data, the cognitive data is resampled using a linear interpolation algorithm to generate time-aligned <blood glucose value, reaction efficacy value> sequences and <blood glucose value, memory efficacy value> sequences. The original reaction test data is in milliseconds, and the original memory test data is in the number of scores.

[0051] Next, based on the above aligned sequences, reaction time-glucose level sub-models and memory-glucose level sub-models were constructed. Gaussian mixture regression algorithm was used to fit the nonlinear effect of glucose on cognition. The sub-model fitting formula is as follows: ,in G is the predicted efficacy value for the k-th cognitive dimension, ranging from 0 to 1, where k represents reaction time or memory, and G is the input blood glucose value in millimoles per liter. Here, represents the peak performance coefficient in this dimension, and is a dimensionless coefficient. The optimal blood glucose feature point corresponding to this cognitive dimension is expressed in millimoles per liter. The glucose sensitivity bandwidth is expressed in millimoles per liter. The parameters are solved iteratively using the least squares method to quantify the bell-shaped curve distribution characteristics of a single cognitive function as a function of blood glucose, thereby generating reaction-glucose sub-model data and memory-glucose sub-model data.

[0052] Subsequently, imaging data of the patient's cerebrovascular lesions were acquired. By registering the imaging data with standard brain atlases, the spatial distribution of the lesions in brain functional areas was determined, and the regulatory weight of the brain functional areas where the lesions were located on reaction time and memory was calculated. This weight calculation was based on a voxel-level lesion assessment model, and the weight calculation formula is as follows: ,in The regulatory weights for the k-th cognitive dimension represent the remaining functional integrity of this brain region. The total volume of the brain region responsible for type k cognitive function in a standard brain atlas, sourced from a medical atlas database; The overlap volume between the lesion region and the corresponding functional brain region is extracted using Boolean operations; This is the neural compensation coefficient, a dimensionless coefficient with a default value of 1. It is dynamically adjusted upwards if increased activity in the surrounding area is detected.

[0053] Finally, based on the regulatory weights of brain functional areas on reaction time and memory, the reaction time-glucose sub-model and the memory-glucose sub-model are weighted and fused to generate a glucose cognitive efficacy correlation model. The fusion calculation formula is as follows: ,in To comprehensively assess the correlation between blood glucose and cognitive efficacy, and These are the regulatory weights for reaction time and memory, respectively. and The output values ​​of the reaction ability and memory sub-models are respectively used. The bias of cognitive test data caused by brain structural damage is eliminated by weighted averaging, so as to output blood glucose cognitive efficacy correlation model data that can truly reflect the impact of blood glucose on the overall brain operating efficiency under the patient's current physiological state.

[0054] In the above method, based on the peak distribution of executive function scores in the model, the functional blood glucose threshold range of the patient is defined in reverse, including:

[0055] In the blood glucose cognitive efficacy association model, the region where the executive function score reaches a preset percentage above the historical maximum value is identified as the optimal cognitive efficacy region;

[0056] Extract all blood glucose values ​​that fall within the optimal cognitive efficiency zone, and determine the confidence interval of the blood glucose values ​​by fitting a normal distribution.

[0057] The lower and upper limits of the confidence interval are set as the functional hypoglycemia threshold and the functional hyperglycemia threshold, respectively, thus forming the functional blood glucose threshold interval.

[0058] Specifically, the optimal cognitive efficiency zone is screened and data extracted based on a blood glucose cognitive efficiency association model. This process employs a peak relative threshold truncation method, with the core screening formula being: ,in Let g be the set of blood glucose values ​​corresponding to optimal performance, and g be any blood glucose sampling point within the model's domain. This is the output value of the blood glucose cognitive efficacy association model. This represents the theoretical peak of the patient's cognitive efficacy in historical data. The efficiency cutoff coefficient is 0.85, a dimensionless coefficient, set according to the principle of statistical significance. Next, the extracted blood glucose value set... A normal distribution is fitted to the probability density function to determine the statistical distribution characteristics of the patient's optimal glycemic state. The distribution parameters are calculated using the following formula: as well as ,in The optimal blood glucose expectation, expressed in mmol / L, represents the central blood glucose level at which the patient's cognitive function is most active. The optimal blood glucose standard deviation represents the degree of blood glucose dispersion allowed to maintain a high-efficiency state; N is the set... The number of samples in For the i-th blood glucose value in the set, this process transforms discrete blood glucose points into a continuous probability distribution model through maximum likelihood estimation.

[0059] Finally, the confidence interval for blood glucose values ​​is calculated based on the fitted distribution parameters, thereby defining the functional blood glucose threshold range. The calculation formula is as follows: ,in The functional hypoglycemia threshold, For functional hyperglycemia threshold, The confidence level coefficients ultimately form a set of confidence level coefficients. Functional blood glucose threshold range data with boundary information.

[0060] It should be added that the method also includes a dynamic calibration step for the functional glucose threshold range:

[0061] After generating a health intervention early warning signal, we continuously track the patient's executive function score feedback after adopting energy compensation recommendations;

[0062] If the functional blood glucose score fails to recover to the historical average level within a preset time, the current functional blood glucose threshold range is deemed invalid.

[0063] Trigger a refactoring instruction for the blood glucose cognitive efficacy association model, and recalculate and narrow the range of functional blood glucose thresholds using the latest review data.

[0064] Regression analysis was performed on continuous blood glucose monitoring data and cognitive task load coefficients to calculate the cognitive glucose consumption decay rate under different cognitive loads.

[0065] The functional blood glucose threshold range refers to a specific blood glucose range determined based on a blood glucose cognitive efficacy correlation model to ensure efficient brain function. Specifically, it can be generated using a statistical confidence interval algorithm based on the blood glucose distribution corresponding to cognitive peaks, serving as a trigger boundary for health intervention warnings. The executive function score feedback refers to the real-time quantitative score sequence exhibited by the patient in a mini-cognitive game after adopting energy compensation suggestions. This can be achieved by real-time collection and standardization of interaction reaction time and task accuracy data, used to assess the actual recovery effect of intervention measures on cognitive efficacy. The historical mean level refers to the long-term average cognitive score benchmark exhibited by the patient under physiological homeostasis conditions. This can be constructed by using a moving average or cluster center extraction on multiple sets of historical review data, serving as a quantitative reference for determining whether the current threshold range has become invalid. The model reconstruction instruction refers to the control signal that triggers the system to relearn and correct the mapping relationship between blood glucose and cognition. Specifically, it can be automatically generated using an event-driven mechanism when efficacy recovery is detected as insufficient, initiating a parameter adaptive adjustment process. The shrinking of the functional blood glucose threshold range refers to reducing the upper and lower limits of blood glucose control fluctuations based on the latest review data. This can be achieved using optimization algorithms that reduce the confidence coefficient of the normal distribution or increase the weight of the penalty term, thereby improving the precision of defense against cognitive impairment risk through a more stringent blood glucose control range. Regression analysis is the statistical calculation process that establishes the mathematical relationship between the cognitive task load coefficient and the rate of change in blood glucose. This can be achieved using multinomial regression or least squares fitting, revealing the physiological energy consumption characteristics under different brain computational intensities. The cognitive glucose decay rate characterizes the amount of blood glucose prediction decrease caused by a specific cognitive load per unit time. This can be obtained by calculating the partial derivative of the correlation between the slope of blood glucose change over time and the cognitive load coefficient, providing core calculation parameters for subsequent blood glucose trajectory prediction and early intervention.

[0066] The working process and principle of this application are as follows: First, regression analysis is performed on continuous blood glucose monitoring data and cognitive task load coefficients to calculate the cognitive glucose decay rate under different cognitive loads. After generating a health intervention early warning signal, the system continuously tracks the patient's executive function score feedback after adopting energy compensation suggestions. Next, the system determines whether the executive function score has recovered to the historical average level within a preset time. If it fails to recover, the current functional blood glucose threshold range is deemed invalid, and a reconstruction instruction for the blood glucose cognitive efficacy correlation model is triggered. Finally, the latest review data is used to recalculate and narrow the range of the functional blood glucose threshold range, further focusing the blood glucose control target towards the cognitive efficacy optimal region.

[0067] In the above method, regression analysis is performed on continuous blood glucose monitoring data and cognitive task load coefficients to calculate the cognitive glucose depletion rate under different cognitive loads, including:

[0068] High-intensity task time periods with cognitive task load coefficients greater than a preset high load threshold were selected from historical review datasets.

[0069] Calculate the slope of the decrease in blood glucose value over time in continuous blood glucose monitoring data during a high-intensity task period to obtain the instantaneous decay rate;

[0070] Establish a mapping table between cognitive task load coefficient and instantaneous decay rate, and generate a cognitive glucose decay rate for the patient based on the mapping table.

[0071] Specifically, an adaptive selection of high-intensity task intervals is performed based on the cognitive task load coefficient sequence in the historical review dataset. The selection logic formula is as follows: ,in Let t be the set of high-intensity task time periods, t be the sampling time, and L(t) be the normalized cognitive task load coefficient. The high-load threshold is typically set to 0.7 to cover the top 30% of high-intensity mental activities. The duration of this load, in minutes. The minimum effective duration is set to 5 minutes to eliminate brief pulse interference.

[0072] Next, for each selected high-intensity task time period, the instantaneous blood glucose decay rate was calculated from the synchronized continuous blood glucose monitoring data. A least squares linear regression model was used to extract the blood glucose decline trend. The calculation formula is as follows: ,in This represents the instantaneous decay rate, measured in millimoles per liter per minute. The negative sign is used to convert the decay rate with a decreasing slope into a positive value. 'n' represents the number of data points within this time period. For the time point of the i-th sampling point, The average over time. Let i be the blood glucose value at the i-th sampling point. The calculation process eliminates single-point measurement noise and quantifies the net rate of blood glucose consumption under a specific load, using the mean blood glucose level.

[0073] Finally, a mapping table between the cognitive task load coefficient and the instantaneous decay rate was established, and a cognitive glucose consumption decay rate model was generated. A weighted polynomial fitting algorithm was used, and the core fitting formula is as follows: ,in The predicted cognitive glucose depletion rate for a specific load L is expressed in mmol / L / min. The cognitive metabolic sensitivity coefficient, expressed in mmol / L / min, characterizes the rate of additional glucose consumption caused by a unit increase in cognitive load, and is derived through regression analysis; L represents the input cognitive task load coefficient. It is the basal metabolic glucose consumption rate, expressed in mmol / L / min, which characterizes the natural decline of blood glucose in patients at rest or under low load. The residual correction term, in units of mmol / L / min, is dynamically compensated based on the moving average of historical prediction errors, ultimately generating a quantitative model that can output the corresponding blood glucose consumption rate according to different cognitive task intensities.

[0074] Fluctuation lag analysis was performed on historical review datasets to determine the impairment lag time parameter that caused the decline in executive function scores due to abnormal fluctuations in blood glucose levels.

[0075] In the above method, the impairment lag time parameter for determining the decline in executive function scores due to abnormal fluctuations in blood glucose levels includes:

[0076] Identify the starting point of abnormal fluctuations in continuous blood glucose monitoring data that deviate from the functional blood glucose threshold range;

[0077] In the time series following the onset of abnormal fluctuations, the first derivative of the performance score value over time is calculated; the point at which the first derivative first falls below the preset performance decay gradient threshold is marked as the performance inflection point.

[0078] The time difference between the performance inflection point and the onset of abnormal fluctuations is calculated, and the average of multiple time differences is determined as the damage lag time parameter.

[0079] Specifically, it involves traversing continuous blood glucose monitoring data sequences based on functional blood glucose threshold ranges. Anomaly events are located using a boundary crossing detection algorithm. The starting point determination formula is as follows: ,in Let k be the starting time of the kth abnormal fluctuation. This represents the time when the last abnormal event ended, and its initial value is 0. Let t be the real-time blood glucose value at time t. and These are the thresholds for functional hypoglycemia and hyperglycemia, respectively. Then, at the starting point of each abnormal fluctuation... In the subsequent time series, a sliding search of the performance score values ​​is performed to locate the performance inflection point. This process uses a sliding window gradient detection algorithm, and the core calculation formula is as follows: ,in This represents the inflection point of efficiency decline corresponding to the k-th fluctuation. W represents the time variable for the sliding search, and W is the width of the sliding window, measured in minutes. It is typically set to 5 minutes to smooth out instantaneous measurement noise. for The performance score at any given time is a normalized dimensionless value. The threshold for a significant decrease in gradient is a dimensionless threshold, set to -0.05 based on the variance of historical data, indicating that cognitive efficacy has begun to show a statistically significant decline.

[0080] Finally, the time difference between the performance inflection point and the onset of abnormal fluctuations is calculated, and the average of multiple time differences is determined as the damage lag time parameter. The calculation formula is as follows: ,in The final determined damage lag time parameter, in minutes, is N, where N is the total number of valid fluctuation events detected within the review period. For the response time delay of a single event, Here, is the weighting coefficient for the k-th event, and is a dimensionless coefficient, calculated based on . That is, the peak value of the fluctuation. Deviation from the nearest boundary The amplitude and width of the normal blood glucose range The ratio of blood sugar fluctuations is calculated based on the logic that more dramatic blood sugar fluctuations contribute a greater weight to the parameter, resulting in a higher final output. It is a physiological response lag index that comprehensively considers the weight of fluctuation intensity.

[0081] Real-time monitoring of patients' current blood glucose levels and the real-time load coefficient of ongoing cognitive tasks allows for the prediction of future blood glucose trajectory using cognitive glucose decay rate.

[0082] The above methods utilize cognitive glucose depletion rate to predict future blood glucose trajectory, including:

[0083] Get the current real-time blood glucose value and the current real-time load factor ;

[0084] Search the mapping table for... Matching cognitive glucose decay rate ;

[0085] Based on formula Simulate and calculate the predicted blood glucose value within a preset time window t, and generate a blood glucose change trajectory;

[0086] The above This represents a predicted blood glucose level for the patient at a future time point t. This blood glucose level is based on the current real-time blood glucose level. and the real-time load coefficient of the current cognitive task By calculating the rate of decline in blood glucose over a future period (cognitive glucose decay rate) This calculation reflects the impact of cognitive activity on glucose consumption, thus helping to predict when intervention is needed to prevent blood glucose from falling below the functional threshold.

[0087] Specifically, the system bus reads the current real-time blood glucose value uploaded by the continuous glucose monitoring sensor. And the current real-time load coefficient output by the micro-cognitive game backend. At the same time, set the prediction time window length. The timeframe is typically set to 30 to 60 minutes to cover short-term metabolic effects, and is based on a mapping table of cognitive task load coefficients and instantaneous decay rates. The corresponding cognitive glucose depletion rate was found using linear interpolation. The unit is millimoles per liter per minute (mmol / L), which represents the expected net blood glucose consumption per minute under the current high-intensity mental workload. Next, based on the acquired initial state parameters, a simulation of the future blood glucose evolution trajectory in the time domain is performed. This process employs a discretized numerical integration algorithm, and the core prediction model formula is: In digital signal processing, this is specifically manifested as ,in To predict the blood glucose level of patients at a future time point t, ; To predict the baseline blood glucose value at the start time, The simulation time step is set to 1 minute to ensure resolution; For the integration variable The decay rate corresponding to the cognitive load projected at any given time, expressed in mmol / L / min, is taken as a value obtained from a table, assuming the patient continues performing the current intensity cognitive task. The integral term The physical meaning is the total amount of blood glucose consumed due to high-intensity cognitive activity within a future time period t. This calculation process integrates the instantaneous metabolic rate over time to derive a dynamic curve of blood glucose concentration that decreases monotonically over time under the condition of no external sugar supplementation and maintaining the current mental load, ultimately generating a blood glucose change trajectory.

[0088] If the predicted blood glucose trajectory falls outside the functional blood glucose threshold range before the damage lag time parameter is reached, a health intervention warning signal containing energy compensation recommendations is generated.

[0089] In the above method, the logic for generating energy compensation recommendations in the health intervention early warning signal is as follows:

[0090] Based on the predicted blood glucose change trajectory, calculate the predicted time when blood glucose levels fall below the functional blood glucose threshold range.

[0091] Calculate the difference between the predicted time of boundary breach and the current system time, and define this difference as the remaining safety buffer duration;

[0092] A reverse compensation model based on cognitive dynamics is constructed. The remaining safe buffer time and the cognitive glucose consumption decay rate are substituted into the model to calculate the estimated amount of carbohydrates required to maintain cognitive efficacy within the remaining safe buffer time. The estimated amount of carbohydrates is then encapsulated into a health intervention early warning signal.

[0093] Specifically, the core calculation formula for determining the critical drop point based on the blood glucose change trajectory is as follows: ,in This refers to the critical time point when blood glucose levels drop below the functional threshold range. To predict the window length, To predict blood glucose levels on the trajectory.

[0094] Next, the difference between the critical time point and the current time is calculated to assess the level of urgency. The calculation formula is as follows: ,in For the remaining safety buffer time, The current system time, if Less than the preset replenishment response threshold If this is detected, an imminent risk of cognitive impairment is identified, immediately triggering a health intervention warning signal. This represents the physical delay in acquiring food and beginning digestion, which can be set to 15 minutes. Finally, while generating the signal, the estimated amount of carbohydrates to be ingested within the remaining safe buffer time is calculated. This calculation employs a cognitive dynamics-based inverse compensation model, with the core formula being: ,in This is the estimated recommended intake of carbohydrates, in grams. The expected blood glucose level corresponding to optimal cognitive performance. This is the lower threshold that is about to be breached. The cognitive glucose depletion rate under the current cognitive load. The load duration weighting coefficient is a dimensionless coefficient, usually taken as 1.0-1.2, used to compensate for the extra consumption caused by sustained high mental activity during digestion; This delays the onset of action of carbohydrates on average. The carbohydrate sensitivity coefficient for patients is expressed in mmol / L / g, which represents the increase in blood glucose concentration caused by ingesting 1 gram of carbohydrates.

[0095] In the above method, if the predicted blood glucose trajectory falls outside the functional blood glucose threshold range before the damage lag time parameter is reached, a health intervention warning signal containing energy compensation suggestions is generated, specifically including:

[0096] Based on the aforementioned damage lag time parameter, the theoretical moment of cognitive impairment under the current high cognitive load state is deduced;

[0097] Obtain the predicted out-of-bounds moment and establish time series comparison logic;

[0098] When the predicted time of exceeding the limit is determined to be earlier than the theoretical time of cognitive impairment, it is identified as a latent risk of glucose deficiency, and a health intervention warning signal is immediately triggered based on the predicted time of exceeding the limit.

[0099] Specifically, in the process of generating a health intervention warning signal containing energy compensation recommendations if the predicted blood glucose trajectory falls outside the functional blood glucose threshold range before the damage lag time parameter is reached, the first step is to perform precise temporal location of the predicted boundary crossing time. This process is based on the blood glucose trajectory. Lower limit of functional blood glucose threshold range Numerical calculations are performed, and the trajectory intersection algorithm is used to determine the predicted boundary crossing time. The calculation formula is as follows: ,in To predict the time of boundary crossing, the unit is minutes, representing the time relative to the current moment. The relative time.

[0100] Next, based on the damage lag time parameter, the theoretical moment of cognitive impairment is deduced, and a cognitive impairment time series model is established. The calculation formula is as follows: ,in The predicted time of cognitive impairment is expressed in minutes. This is the starting point for detecting current high-load cognitive tasks or abnormal fluctuation trends. The damage lag time parameter characterizes the compensatory buffering capacity of the patient's brain in the early stages of glucose deficiency. The current real-time load correction factor is a dimensionless parameter with a value range of 0-1. It is obtained by normalizing the current cognitive task load coefficient and is used to reflect the shortening effect of compensation time under high load.

[0101] Finally, the timing comparison and early warning triggering logic is executed. A health intervention early warning signal is generated by comparing the temporal relationship between the predicted boundary violation time and the cognitive impairment time. The determination logic formula is as follows: ,in As a warning trigger indicator, and All are time values ​​relative to the same reference point. < At the time of its establishment, it was indicated that biochemical indicators of blood sugar would fall below the safe threshold before brain function completely failed, and would... The optimal early warning trigger point is established to generate a health intervention early warning signal aimed at blocking the subsequent decline in cognitive efficacy. This signal provides energy replenishment before cognitive impairment occurs, thus achieving a protective intervention for the brain function of patients with cerebrovascular disease.

[0102] It should be added that if the predicted blood glucose trajectory falls outside the functional blood glucose threshold range after the damage lag time parameter is reached, a cognitively driven defensive intervention mechanism is activated, generating a preventative warning signal that includes suggestions for dynamic threshold elevation, specifically including:

[0103] First, when comparing and determining the predicted boundary crossing time. Later than or equal to the moment of cognitive impairment Subsequently, the system identifies a latent glucose deficiency state characterized by "high cognitive consumption and low glycemic sensitivity," and then calculates the "functional threshold drift" caused by accumulated cognitive load. This calculation employs a time-metabolic coupling compensation algorithm, with the core formula being: ,in This is the threshold drift compensation amount, expressed in millimoles per liter. The current cognitive glucose depletion rate is expressed in mmol / L / min. The time difference between physical boundary violation and functional impairment. The safety redundancy coefficient is a dimensionless coefficient, typically taken as 1.2. This formula quantifies the additional glucose reserve required to protect cognitive function from decline before blood glucose alarms.

[0104] Next, the "dynamic safety intervention threshold" is calculated based on the drift amount. The calculation formula is as follows: ,in The revised dynamic safety intervention threshold is expressed in mmol / L. This calculation actually dynamically adjusts the static blood glucose baseline based on the current level of awareness and urgency.

[0105] Finally, at the moment of cognitive impairment The warning trigger time is calculated backwards from the baseline, and the energy replenishment requirement is recalculated to generate a preventative warning signal. The replenishment amount is calculated using the following formula: in Recommended intake of carbohydrates for prevention, in grams. The predicted blood glucose level at the time of cognitive impairment is expressed in mmol / L. This is the food absorption time constant, measured in minutes. The continuous load factor is a dimensionless factor. The carbohydrate sensitivity, measured in mmol / L / g, is a signal that, by forcibly raising the intervention standard, ensures that energy is replenished in advance when the blood glucose level has not yet reached its lowest point but the brain is about to "shut down," effectively avoiding accidental damage to cognitive function caused by mechanically applying a fixed threshold.

[0106] Among them, the continuous load factor This is used to assess the impact of sustained cognitive tasks on glycemic expenditure. The numerical value can help adjust recommended intake to better compensate for the additional expenditure under prolonged high-load conditions. It is typically determined through analysis of historical cognitive load data. Statistical models can be used to categorize different cognitive tasks, calculate the relative long-term impact of each task on glycemic expenditure, and then normalize the results.

[0107] Carbohydrate sensitivity This characterizes the degree to which ingested carbohydrates raise blood glucose levels. The body's response to carbohydrate intake may vary among different populations. It is typically determined through regression analysis of individual dietary and blood glucose variability records. Continuous glucose monitoring systems, combined with detailed dietary records, can provide high-precision datasets for model fitting—such as linear regression or multivariate analysis—to obtain individualized sensitivity parameters.

[0108] Food absorption time constant This time constant is used to characterize the time delay in reaching the peak glycemic effect after carbohydrate ingestion. Taking into account the type of food and the rate of digestion and absorption, this time constant can significantly affect the effectiveness of timely replenishment. This constant is usually obtained through physiological experiments or literature studies. The average absorption time can be determined by experimentally monitoring different types of carbohydrates.

[0109] Example 2:

[0110] like Figure 3 As shown, the chronic disease health management device based on key indicator correlation analysis includes a processor, memory, and communication bus.

[0111] The memory stores a computer-readable program that can be executed by the processor;

[0112] The communication bus enables communication between the processor and the memory;

[0113] When the processor executes the computer-readable program, it performs steps in the chronic disease health management method based on key indicator correlation analysis as described in any one of the present invention.

[0114] This invention uses a technique based on the optimal operational efficiency of the individual brain to define the normal blood glucose range in reverse. This cognitive performance-oriented dynamic calibration method can accurately capture individual performance differences at different blood glucose levels, thereby defining a blood glucose range that better reflects actual needs and providing a scientific basis for personalized health management. This feature ensures that patients with diabetes and cerebrovascular disease can maintain optimal functional status under complex cognitive tasks, significantly reducing errors and interference from traditional monitoring.

[0115] Unlike traditional systems, this invention not only monitors the impact of blood glucose on cognition in real time, but also introduces an estimation and early warning mechanism for blood glucose consumption triggered by high-intensity cognitive activity. Through the simultaneous acquisition and analysis of continuous blood glucose monitoring data and cognitive performance, the system can predict potential hypoglycemia risks in the early stages of increased cognitive load and issue early glucose supplementation recommendations. This predictive intervention mechanism significantly improves the initiative and timeliness of monitoring, strengthens the dynamic response to the brain's energy demands, and thus ensures the stability of the patient's cognitive abilities under constantly fluctuating physiological conditions.

[0116] The innovation of this technical solution lies not only in proposing a completely new monitoring indicator system, but also in the successful design of a blood glucose threshold framework that can dynamically adjust its upper and lower limits. By integrating data on blood glucose changes and cognitive performance, the system complements and innovates traditional blood glucose monitoring in multiple dimensions, promoting the development of personalized precision medicine.

[0117] It should be noted that the interval and threshold sizes are set for ease of comparison. The size of the threshold depends on the amount of sample data and the base number set by those skilled in the art for each set of sample data, as long as it does not affect the proportional relationship between the parameter and the quantized value. Furthermore, the above formulas are all dimensionless calculations, and the formulas are derived from software simulations using a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0118] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0119] It should be understood that determining B based on A does not mean determining B solely based on A; it also means determining B based on A and / or other information.

[0120] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0121] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A chronic disease health management method based on key indicator correlation analysis, characterized in that, The method includes the following steps: Obtain the patient's historical follow-up dataset, which includes: continuous blood glucose monitoring data collected synchronously within the preset follow-up time period and micro cognitive game interaction data. The micro cognitive game interaction data includes performance function scores with time points and corresponding cognitive task load coefficients. We performed time-series alignment and multidimensional mapping on the historical review dataset to construct a blood glucose cognitive efficacy association model. Based on the peak distribution of executive function scores in the model, we back-defined the functional blood glucose threshold range for patients. Regression analysis was performed on continuous blood glucose monitoring data and cognitive task load coefficients to calculate the cognitive glucose consumption decay rate under different cognitive loads. Fluctuation lag analysis was performed on historical review datasets to determine the impairment lag time parameter that caused the decline in executive function scores due to abnormal fluctuations in blood glucose levels. Real-time monitoring of patients' current blood glucose levels and the real-time load coefficient of ongoing cognitive tasks; using cognitive glucose decay rate to predict future blood glucose change trajectories. If the predicted blood glucose trajectory falls outside the functional blood glucose threshold range before the damage lag time parameter is reached, a health intervention warning signal containing energy compensation recommendations is generated.

2. The chronic disease health management method based on key indicator correlation analysis according to claim 1, characterized in that, The cognitive task load coefficient in the micro-cognitive game interaction data was determined in the following way: Acquire the game difficulty level, game operation frequency, and game screen refresh rate of the mini cognitive game; The game difficulty level, game operation frequency, and game screen refresh rate are normalized. The numerical value representing the brain's computational intensity is obtained by weighted summation and used as the cognitive task load coefficient.

3. The chronic disease health management method based on key indicator correlation analysis according to claim 1, characterized in that, A time-series alignment and multidimensional mapping were performed on the historical review dataset to construct a blood glucose cognitive efficacy correlation model, including: Based on the interactive data of the mini cognitive game, reaction score and memory score were extracted according to different cognitive dimensions. Reaction score and memory score were correlated with blood glucose values ​​at the same time to construct reaction-blood glucose sub-models and memory-blood glucose sub-models, respectively. Obtain imaging data of cerebrovascular lesions in patients to determine the functional brain regions where the lesions are located; Based on the regulatory weights of brain functional areas on reaction time and memory, the reaction time-glucose sub-model and the memory-glucose sub-model are weighted and fused to generate a glucose cognitive efficacy correlation model.

4. The chronic disease health management method based on key indicator correlation analysis according to claim 1, characterized in that, Based on the peak distribution of executive function scores in the model, the functional glycemic threshold range for patients is defined in reverse, including: In the blood glucose cognitive efficacy association model, the region where the executive function score reaches a preset percentage above the historical maximum value is identified as the optimal cognitive efficacy region; Extract all blood glucose values ​​that fall within the optimal cognitive efficiency zone, and determine the confidence interval of the blood glucose values ​​by fitting a normal distribution. The lower and upper limits of the confidence interval are set as the functional hypoglycemia threshold and the functional hyperglycemia threshold, respectively, thus forming the functional blood glucose threshold interval.

5. The chronic disease health management method based on key indicator correlation analysis according to claim 1, characterized in that, Regression analysis was performed on continuous glucose monitoring data and cognitive task load coefficients to calculate the cognitive glucose depletion rate under different cognitive loads, including: High-intensity task time periods with cognitive task load coefficients greater than a preset high load threshold were selected from historical review datasets. Calculate the slope of the decrease in blood glucose value over time in continuous blood glucose monitoring data during a high-intensity task period to obtain the instantaneous decay rate; Establish a mapping table between cognitive task load coefficient and instantaneous decay rate, and generate a cognitive glucose decay rate for the patient based on the mapping table.

6. The chronic disease health management method based on key indicator correlation analysis according to claim 5, characterized in that, Predicting future blood glucose trajectories using cognitive glucose decay rate, including: Get the current real-time blood glucose value and the current real-time load factor ; Search the mapping table for... Matching cognitive glucose decay rate ; Based on formula Simulate and calculate the predicted blood glucose value within a preset time window t, and generate a blood glucose change trajectory; The above This represents the predicted blood glucose level of a patient at a future time point t.

7. The chronic disease health management method based on key indicator correlation analysis according to claim 1, characterized in that, The logic for generating energy compensation recommendations in health intervention early warning signals is as follows: Based on the predicted blood glucose change trajectory, calculate the predicted time when blood glucose levels fall below the functional blood glucose threshold range. Calculate the difference between the predicted time of boundary breach and the current system time, and define this difference as the remaining safety buffer duration; A reverse compensation model based on cognitive dynamics is constructed. The remaining safe buffer time and the cognitive glucose consumption decay rate are substituted into the model to calculate the estimated amount of carbohydrates required to maintain cognitive efficacy within the remaining safe buffer time. The estimated amount of carbohydrates is then encapsulated into a health intervention early warning signal.

8. The chronic disease health management method based on key indicator correlation analysis according to claim 1, characterized in that, Determine the impairment lag time parameters for the decline in executive function scores caused by abnormal fluctuations in blood glucose levels, including: Identify the starting point of abnormal fluctuations in continuous blood glucose monitoring data that deviate from the functional blood glucose threshold range; In the time series following the onset of abnormal fluctuations, the first derivative of the performance score value over time is calculated; the point at which the first derivative first falls below the preset performance decay gradient threshold is marked as the performance inflection point. The time difference between the performance inflection point and the onset of abnormal fluctuations is calculated, and the average of multiple time differences is determined as the damage lag time parameter.

9. The chronic disease health management method based on key indicator correlation analysis according to claim 8, characterized in that, If the predicted blood glucose trajectory falls outside the functional blood glucose threshold range before the damage lag time parameter is reached, a health intervention warning signal containing energy compensation recommendations is generated, specifically including: Based on the aforementioned damage lag time parameter, the theoretical moment of cognitive impairment under the current high cognitive load state is deduced; Obtain the predicted out-of-bounds moment and establish time series comparison logic; When the predicted time of exceeding the limit is determined to be earlier than the theoretical time of cognitive impairment, it is identified as a latent risk of glucose deficiency, and a health intervention warning signal is immediately triggered based on the predicted time of exceeding the limit.

10. A chronic disease health management device based on key indicator correlation analysis, characterized in that, Includes processor, memory, and communication bus; The memory stores a computer-readable program that can be executed by the processor; The communication bus enables communication between the processor and the memory; When the processor executes the computer-readable program, it performs the steps in the chronic disease health management method based on key indicator correlation analysis as described in any one of claims 1 to 9.

Images