Diabetic patient clinical data management system and method based on feature extraction

Abstract

The invention relates to the technical field of medical care informatics, in particular to a diabetic clinical data management system and method based on feature extraction, and the method comprises the steps: obtaining patient clinical data of a diabetic, the patient clinical data including patient blood glucose data and patient behavior data; behavior matching features among different behaviors of the diabetic patient are extracted through the patient behavior data; extracting correlation characteristics of different behaviors of the diabetic patient and blood glucose fluctuation through the patient blood glucose data and the patient behavior data; evaluating the blood glucose fluctuation risk according to the behavior matching features and the associated features of the diabetic patient, and performing blood glucose fluctuation risk early warning according to a blood glucose fluctuation risk evaluation result. According to the method, the accuracy and the reliability of blood glucose fluctuation risk early warning are effectively improved by quantifying the behavior matching degree among the dining behavior, the medication behavior and the movement behavior of the diabetic patient.

Description

Technical Field

[0001] The present application relates to the technical field of medical care informatics, and in particular to a clinical data management system and method for diabetic patients based on feature extraction. Background Art

[0002] Diabetes is a chronic disease with symptoms of high blood sugar caused by insufficient insulin or poor cellular response to insulin. Long-term high blood sugar and blood sugar fluctuations are harmful to the patient's body. The daily behavior and self-management ability of diabetic patients are key factors affecting treatment.

[0003] Diet therapy, exercise therapy, drug therapy and self-blood sugar monitoring are important components of diabetes treatment. By reasonably adjusting the diet structure, actively participating in appropriate exercise, taking medications on time and regularly monitoring blood sugar, diabetic patients can better control blood sugar levels, reduce the risk of complications and improve their quality of life.

[0004] In real life, the eating, medication and exercise behaviors of diabetic patients need to match each other. For example, matching eating time with medication time can ensure that the medication can effectively regulate blood sugar levels after meals. When the meal size is reduced, the medication dosage should be reduced accordingly to avoid the risk of hypoglycemia. At the same time, the patient's exercise behavior, eating and medication behavior must also be coordinated to prevent blood sugar from getting out of control due to excessive exercise or insufficient energy supplementation. However, existing technologies ignore the degree of matching between diabetic patients' eating, medication and exercise behaviors when predicting blood sugar fluctuation risks, thereby reducing the prediction accuracy of blood sugar fluctuation risks. Summary of the invention

[0005] In order to overcome the defects and shortcomings of the existing technology, the present application provides a clinical data management system and method for diabetic patients based on feature extraction, which effectively improves the accuracy and reliability of blood sugar fluctuation risk warning by quantifying the degree of behavioral matching between diabetic patients' dining behavior, medication behavior and exercise behavior.

[0006] In order to achieve the above objectives, this application adopts the following technical solutions: In a first aspect, the present application provides a method for managing clinical data of diabetic patients based on feature extraction, comprising the following steps: Acquire clinical data of diabetic patients, where the clinical data includes blood sugar data and behavior data of the patients; Extract behavioral matching features between different behaviors of diabetic patients through patient behavior data; Extract the correlation characteristics between different behaviors of diabetic patients and blood sugar fluctuations through patient blood sugar data and patient behavior data; The blood sugar fluctuation risk is assessed based on the behavioral matching characteristics and associated characteristics of diabetic patients, and blood sugar fluctuation risk warnings are issued based on the blood sugar fluctuation risk assessment results.

[0007] Optionally, the specific steps of extracting behavior matching features between different behaviors of diabetic patients include: Acquire patient behavior data, which includes dining behavior data, medication behavior data, and exercise behavior data of diabetic patients; Extracting behavior matching features between different behaviors through patient behavior data, the behavior matching features include first behavior matching features, second behavior matching features and third behavior matching features; Among them, the first behavior matching feature is used to evaluate the matching degree between the dining behavior and medication behavior of diabetic patients, and the first behavior matching feature is calculated through the dining behavior data and medication behavior data; The second behavior matching feature is used to evaluate the matching degree between the dining behavior and the exercise behavior of the diabetic patients. The second behavior matching feature is calculated through the dining behavior data and the exercise behavior data. The third behavior matching feature is used to evaluate the matching degree between the medication behavior and exercise behavior of diabetic patients. The third behavior matching feature is calculated through the medication behavior data and the exercise behavior data.

[0008] Optionally, the specific steps of calculating the first behavior matching feature through the dining behavior data and the medication behavior data include: Obtaining the patient's dining behavior data and medication behavior data from the patient's behavioral data, wherein the dining behavior data includes dining time, meal weight, and meal food type, and the medication behavior data includes medication time, medication dosage, and medication type; The difference between the meal time and the medication time is taken as the medication time interval, and the medication time matching degree is obtained by comparing the medication time interval with the optimal medication time interval corresponding to the drug type. When the medication time interval is less than or equal to the optimal medication time interval, the medication time matching degree is 1, and when the medication time interval is greater than the optimal medication time interval, the medication time matching degree is 0; The carbohydrate intake is determined by the weight and type of meal food, and the dosage matching degree is calculated by the carbohydrate intake content and the standard carbohydrate content corresponding to the dosage. The dosage matching degree calculation formula is as follows: ; In the formula Indicates the matching degree of drug dosage. Indicates carbohydrate intake. Indicates the standard carbohydrate content corresponding to the dosage; The mean of the medication time matching degree and the medication dosage matching degree is taken as the first row matching feature.

[0009] Optionally, the specific steps of calculating the second behavior matching feature through the dining behavior data and the exercise behavior data include: Acquire dining behavior data and exercise behavior data, wherein the dining behavior data includes dining time, meal weight and meal type, and the exercise behavior data includes exercise start time and exercise time set; The difference between the meal time and the exercise start time is taken as the exercise time interval, and the second behavior matching feature is obtained by comparing the exercise time interval with the recommended exercise time interval. When the exercise time interval is greater than the recommended exercise time interval, the second behavior matching feature is 1, and when the exercise time interval is less than or equal to the exercise time interval, the second behavior matching feature is 0.

[0010] Optionally, the specific steps of calculating the third behavior matching feature through the medication behavior data and the exercise behavior data include: Obtain medication behavior data and exercise behavior data, the medication behavior data includes medication time, medication dosage and medication type, and the exercise behavior data includes exercise start time and exercise time set; Obtain the maximum drug effect time set by the maximum drug effect time corresponding to the medication time and drug type, and calculate the number of elements of the intersection of the movement time set and the maximum drug effect time set and the number of elements of the union of the movement time set and the maximum drug effect time set; Calculate the matching features of the third line by the number of intersection elements and the number of union elements , ,in, Represents the number of elements in the intersection of the exercise time set and the maximum drug effect time set, Represents the number of elements in the union of the exercise time set and the maximum drug effect time set, Indicates that the third line is a matching feature.

[0011] Optionally, the specific steps of extracting the correlation features between different behaviors of diabetic patients and blood sugar fluctuations include: Obtaining patient blood sugar data and patient behavior data, wherein the patient behavior data includes the dining behavior data, medication behavior data and exercise behavior data of diabetic patients; Extracting behavioral time points through patient behavioral data, the behavioral time points include the patient's meal behavior time points, medication behavior time points, and exercise behavior time points; The blood sugar fluctuation time points are extracted from the patient's blood sugar data and the correlation characteristics between each behavior type and blood sugar fluctuation are calculated in combination with the patient's behavior time points. The correlation characteristic calculation formula is as follows: ; In the formula Indicates patient behavior No. Behavior time point, Indicates the patient behavior type, represent eating behavior, medication behavior, and exercise behavior, respectively. Indicates patient behavior The number of behavioral time points, Indicates Blood sugar fluctuation time point, Indicates the number of blood sugar fluctuation time points, Indicates patient behavior No. The time difference between a behavior time point and the most recent blood sugar fluctuation after the behavior time point, Indicates the data monitoring cycle, Indicates patient behavior Characteristics associated with blood sugar fluctuations.

[0012] Optionally, the specific steps of assessing the blood sugar fluctuation risk and providing a blood sugar fluctuation risk warning according to the blood sugar fluctuation risk assessment result include: Obtaining behavioral matching features and correlation features between different behaviors and blood sugar fluctuations; The blood sugar fluctuation risk index is calculated through the behavior matching characteristics and the association characteristics. The blood sugar fluctuation risk index is used to assess the patient's blood sugar fluctuation risk. The blood sugar fluctuation risk index calculation formula is as follows: ; In the formula Indicates that the first line is a matching feature. Indicates that the second line is a matching feature. Indicates that the third line is a matching feature. Indicates the correlation characteristics between patients' meal behavior and blood sugar fluctuations, Indicates the correlation characteristics between patients' medication behavior and blood sugar fluctuations, Indicates the correlation characteristics between patients' exercise behavior and blood sugar fluctuations, represents the blood sugar fluctuation risk index; When the blood sugar fluctuation risk index is greater than the preset blood sugar fluctuation risk threshold, a blood sugar fluctuation risk warning is issued. When the blood sugar fluctuation risk index is less than or equal to the preset blood sugar fluctuation risk threshold, no blood sugar fluctuation risk warning is issued.

[0013] It should be noted here that the preset blood sugar fluctuation risk threshold is determined as follows: 1,000 groups of patient clinical data are collected to distinguish whether the patient's blood sugar fluctuation risk meets health requirements, the patient's clinical data is substituted into the blood sugar fluctuation risk index calculation formula for calculation, the calculated blood sugar fluctuation risk index and the distinction result are simultaneously imported into the fitting software, and the optimal preset blood sugar fluctuation risk threshold that meets the distinction accuracy of the distinction result is output.

[0014] In a second aspect, the present application provides a clinical data management system for diabetic patients based on feature extraction, comprising: A data acquisition module is used to acquire clinical data of diabetic patients, where the clinical data includes blood sugar data and behavior data of patients; A first feature extraction module is used to extract behavior matching features between different behaviors of diabetic patients through patient behavior data; The second feature extraction module is used to extract the correlation features between different behaviors of diabetic patients and blood sugar fluctuations through the patient blood sugar data and patient behavior data; The risk warning module is used to evaluate the blood sugar fluctuation risk based on the behavioral matching characteristics and associated characteristics of diabetic patients and to issue blood sugar fluctuation risk warning based on the blood sugar fluctuation risk assessment results.

[0015] In a third aspect, the present application provides an electronic device comprising: a processor and a memory, wherein the memory stores a computer program that can be called by the processor, and the processor executes a clinical data management method for diabetic patients based on feature extraction by calling the computer program stored in the memory.

[0016] In a fourth aspect, the present application provides a computer-readable storage medium storing instructions, which, when executed on a computer, enable the computer to execute a clinical data management method for diabetic patients based on feature extraction.

[0017] Compared with the prior art, this application has the following advantages and beneficial effects: The present application extracts the behavioral matching features between the dining behavior, medication behavior and exercise behavior of diabetic patients through patient behavior data, and respectively extracts the correlation features between the dining behavior, medication behavior and exercise behavior of diabetic patients and blood sugar fluctuations through patient blood sugar data and patient behavior data, accurately quantifies the correlation between different behaviors and blood sugar fluctuations, evaluates the blood sugar fluctuation risk of diabetic patients by combining the behavioral matching features and the correlation features, and issues blood sugar fluctuation risk warnings based on the blood sugar fluctuation risk assessment results, effectively improving the accuracy and reliability of blood sugar fluctuation risk warnings. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Other features, objects and advantages of the present application will become more apparent by reading the detailed description of non-limiting embodiments made with reference to the following drawings: Figure 1 It is a schematic diagram of the overall process of the clinical data management method for diabetic patients based on feature extraction provided in an embodiment of the present application; Figure 2 It is a flow chart of the associated features in the method for managing clinical data of diabetic patients based on feature extraction provided in an embodiment of the present application; Figure 3 It is a structural diagram of a clinical data management system for diabetic patients based on feature extraction provided in an embodiment of the present application; Figure 4 It is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0019] The technical solution of the present application is described in detail below through the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present application and the specific features in the embodiments are detailed descriptions of the technical solution of the present application, rather than limitations on the technical solution of the present application. In the absence of conflict, the embodiments of the present application and the technical features in the embodiments can be combined with each other.

[0020] See also Figure 1 , Figure 1 The overall process diagram of the method for managing clinical data of diabetic patients based on feature extraction provided in the embodiment of the present application specifically includes the following steps: S110: Acquire clinical data of the diabetic patient, where the clinical data includes blood sugar data and behavior data of the patient.

[0021] S120: Extracting behavioral matching features between different behaviors of diabetic patients through patient behavior data; The behavioral matching between the dining behavior, medication behavior and exercise behavior of diabetic patients refers to the degree of coordination of these behaviors in terms of timing and intensity in order to achieve stable blood sugar control. The meaning of behavioral matching is that dining behavior needs to provide energy sources for the body, medication behavior should exert the effect of controlling blood sugar at the appropriate time, and exercise behavior assists blood sugar regulation by consuming glucose. When the above behaviors match each other in terms of time, amount and effect, blood sugar fluctuations can be minimized and the risk of hypoglycemia or hyperglycemia can be avoided. The specific steps for extracting the behavioral matching characteristics between different behaviors of diabetic patients include: Acquire patient behavior data, which includes dining behavior data, medication behavior data, and exercise behavior data of diabetic patients; Extracting behavior matching features between different behaviors through patient behavior data, the behavior matching features include first behavior matching features, second behavior matching features and third behavior matching features; Among them, the first behavior matching feature is used to evaluate the matching degree between the dining behavior and medication behavior of diabetic patients, and the first behavior matching feature is calculated through the dining behavior data and medication behavior data; The second behavior matching feature is used to evaluate the matching degree between the dining behavior and the exercise behavior of the diabetic patients. The second behavior matching feature is calculated through the dining behavior data and the exercise behavior data. The third behavior matching feature is used to evaluate the matching degree between the medication behavior and exercise behavior of diabetic patients. The third behavior matching feature is calculated through the medication behavior data and exercise behavior data. Food contains carbohydrates, proteins, fats, etc. Carbohydrates usually refer to starch or sugar contained in food, which is the direct substance that causes blood sugar levels to rise. Good blood sugar control also depends on the speed at which food raises blood sugar. Among the emptying speeds of food, carbohydrates are the fastest, fats are the slowest, and proteins are in between. Therefore, carbohydrates have the greatest impact on postprandial blood sugar. The specific steps for calculating the first behavior matching features through meal behavior data and medication behavior data include: Obtaining the patient's dining behavior data and medication behavior data from the patient's behavioral data, wherein the dining behavior data includes dining time, meal weight, and meal food type, and the medication behavior data includes medication time, medication dosage, and medication type; The difference between the meal time and the medication time is taken as the medication time interval, and the medication time matching degree is obtained by comparing the medication time interval with the optimal medication time interval corresponding to the medication type. When the medication time interval is less than or equal to the optimal medication time interval, the medication time matching degree is 1, and when the medication time interval is greater than the optimal medication time interval, the medication time matching degree is 0, wherein the optimal medication time interval corresponding to the medication type is obtained from the medication instructions; Determine the carbohydrate intake by the weight and type of food consumed. If the type of food consumed is packaged food, refer to the Nutrition Facts Table to obtain the carbohydrate content per 100 grams of food and multiply it by the weight of the food consumed to obtain the carbohydrate intake. If the type of food consumed is weighed food in daily life, refer to the Standard Edition of Chinese Food Composition Table to obtain the carbohydrate content per 100 grams of food and multiply it by the weight of the food consumed to obtain the carbohydrate intake. The dosage matching degree is calculated by the carbohydrate intake content and the standard carbohydrate content corresponding to the dosage. The dosage matching degree calculation formula is as follows: ; In the formula Indicates the matching degree of drug dosage. Indicates carbohydrate intake. Indicates the standard carbohydrate content corresponding to the dosage. It is used to calculate the absolute difference between the carbohydrate intake content and the standard carbohydrate content corresponding to the dosage, that is, to evaluate the deviation between the carbohydrate intake content and the standard carbohydrate content corresponding to the dosage. The standard carbohydrate content in the denominator is For normalization, the standard carbohydrate content corresponding to the dose is calculated by dividing the dose by the insulin-to-carbohydrate ratio (I / Carb); I / Carb is the amount of carbohydrate that can be metabolized by 1 unit of insulin. The size of I / Carb is determined by the patient's insulin sensitivity. The specific steps to determine I / Carb include: the patient records fasting, pre-meal and 2-hour post-meal blood sugar, insulin dose, and carbohydrate intake in meals and snacks for one week. The I / Carb value is obtained by dividing the insulin dose by the total carbohydrate content ingested. For example, if the patient consumes 70 g of carbohydrates and uses 5 units of insulin, and the blood sugar 2 hours after the meal is within the control target range, the I / Carb value is 1:14; The mean of medication time matching and medication dosage matching is taken as the first row matching feature; Matching the dining behavior and exercise behavior of diabetic patients is conducive to achieving the goal of balancing energy supply and demand and stabilizing blood sugar. After a meal, blood sugar usually rises to a peak within a certain period of time. At this time, arranging appropriate exercise can effectively consume blood sugar and avoid postprandial hyperglycemia. If the exercise time is too early or too late, the best time for blood sugar regulation may be missed, and even the risk of hypoglycemia or hyperglycemia may be increased. The specific steps for calculating the second behavior matching feature through dining behavior data and exercise behavior data include: Acquire dining behavior data and exercise behavior data, wherein the dining behavior data includes dining time, meal weight and meal type, and the exercise behavior data includes exercise start time and exercise time set; The difference between the meal time and the exercise start time is taken as the exercise time interval, and the second behavior matching feature is obtained by comparing the exercise time interval with the recommended exercise time interval. When the exercise time interval is greater than the recommended exercise time interval, the second behavior matching feature is 1, and when the exercise time interval is less than or equal to the exercise time interval, the second behavior matching feature is 0. The recommended exercise time interval is obtained from the "Guidelines on Nutrition and Exercise for Hyperglycemia" issued by the National Health Commission of the People's Republic of China, and the recommended exercise time interval defaults to 1 hour; Glucose-lowering drugs (such as insulin and sulfonylureas) will significantly reduce blood sugar levels during the period of maximum drug effect. If you exercise at this time, exercise will further consume glucose, increase the risk of hypoglycemia, and even cause severe hypoglycemia. Therefore, diabetic patients should avoid exercising during the period of maximum drug effect. The specific steps for calculating the third behavior matching feature through medication behavior data and exercise behavior data include: Obtain medication behavior data and exercise behavior data, the medication behavior data includes medication time, medication dosage and medication type, and the exercise behavior data includes exercise start time and exercise time set; Obtain the maximum drug effect time set through the maximum drug effect time corresponding to the medication time and the drug type, calculate the number of elements of the intersection of the movement time set and the maximum drug effect time set, and the number of elements of the union of the movement time set and the maximum drug effect time set, wherein the maximum drug effect time corresponding to the drug type is obtained by referring to the drug instructions; Calculate the matching features of the third line by the number of intersection elements and the number of union elements , ,in, represents the number of elements in the intersection of the exercise time set and the maximum drug effect time set, that is, the length of the overlapping time period between the exercise time and the maximum drug effect time. Represents the number of elements in the union of the exercise time set and the maximum drug effect time set, that is, the total coverage time length of the exercise time and the maximum drug effect time. Indicates that the third line matches the feature. When , it means that the exercise time and drug effect time do not overlap at all, and the matching degree is the highest. When , the exercise time is completely within the maximum drug effect time, and the matching degree is the lowest.

[0022] S130: extracting correlation features between different behaviors of diabetic patients and blood sugar fluctuations through patient blood sugar data and patient behavior data; By extracting the correlation features between different patient behaviors and blood sugar fluctuation time points, we can identify the correlation strength between patient behaviors and blood sugar fluctuations, and thus judge the patient's blood sugar control. For example, the relationship between meal time and blood sugar fluctuations can reveal the risk of postprandial hyperglycemia, the relationship between exercise time and blood sugar fluctuations can help identify the potential risk of hypoglycemia, and the relationship between medication time and blood sugar fluctuations can help judge whether there is a disorder between drug effects and blood sugar control. Please refer to Figure 2 , Figure 2 : is a flow chart of the associated features in the method for managing clinical data of diabetic patients based on feature extraction provided in an embodiment of the present application, and the specific steps include: Obtaining patient blood sugar data and patient behavior data, wherein the patient behavior data includes the dining behavior data, medication behavior data and exercise behavior data of diabetic patients; Extracting behavioral time points through patient behavioral data, the behavioral time points include the patient's meal behavior time points, medication behavior time points, and exercise behavior time points; The blood sugar fluctuation time points are extracted from the patient's blood sugar data and the correlation characteristics between each behavior type and blood sugar fluctuation are calculated in combination with the patient's behavior time points. The correlation characteristic calculation formula is as follows: ; In the formula Indicates patient behavior No. Behavior time point, Indicates the patient behavior type, represent eating behavior, medication behavior, and exercise behavior, respectively. Indicates patient behavior The number of behavioral time points, Indicates Blood sugar fluctuation time point, Indicates the number of blood sugar fluctuation time points, Indicates patient behavior No. The time difference between a behavior time point and the most recent blood sugar fluctuation after the behavior time point, Indicates the data monitoring period, which is used for normalization. The default value of the data monitoring period is 12 hours. Indicates patient behavior Characteristics associated with blood sugar fluctuations.

[0023] S140: Evaluate the risk of blood sugar fluctuation based on the behavioral matching characteristics and associated characteristics of the diabetic patient and issue a blood sugar fluctuation risk warning based on the blood sugar fluctuation risk assessment results; Behavior matching features are used to reveal the coordination between the dining behavior, medication behavior and exercise behavior of diabetic patients. Association features are used to describe the degree of association between different behaviors and blood sugar fluctuations. The behavior matching features and association features are combined to assess whether the patient has the risk of blood sugar fluctuations. The specific steps of assessing the risk of blood sugar fluctuations and issuing blood sugar fluctuation risk warnings based on the blood sugar fluctuation risk assessment results include: Obtaining behavioral matching features and correlation features between different behaviors and blood sugar fluctuations; The blood sugar fluctuation risk index is calculated through the behavior matching characteristics and the association characteristics. The blood sugar fluctuation risk index is used to assess the patient's blood sugar fluctuation risk. The blood sugar fluctuation risk index calculation formula is as follows: ; In the formula Indicates that the first line is a matching feature. Indicates that the second line is a matching feature. Indicates that the third line is a matching feature. Indicates the correlation characteristics between patients' meal behavior and blood sugar fluctuations, Indicates the correlation characteristics between patients' medication behavior and blood sugar fluctuations, Indicates the correlation characteristics between patients' exercise behavior and blood sugar fluctuations, represents the blood sugar fluctuation risk index, where Used to evaluate the matching degree between the dining behavior and medication behavior of diabetic patients. It indicates the joint correlation characteristics of the patient's dining behavior, medication behavior and blood sugar fluctuations. Used to evaluate the matching degree between the dining behavior and exercise behavior of diabetic patients. It indicates the joint correlation characteristics of the patient's dining behavior, exercise behavior and blood sugar fluctuations. Used to evaluate the matching degree between medication behavior and exercise behavior of diabetic patients. It indicates the joint correlation characteristics of the patient's medication behavior, exercise behavior and blood sugar fluctuations; When the blood sugar fluctuation risk index is greater than the preset blood sugar fluctuation risk threshold, a blood sugar fluctuation risk warning is issued. When the blood sugar fluctuation risk index is less than or equal to the preset blood sugar fluctuation risk threshold, no blood sugar fluctuation risk warning is issued.

[0024] See also Figure 3 , Figure 3 : is a structural diagram of a diabetic patient clinical data management system based on feature extraction provided in an embodiment of the present application. This embodiment provides a diabetic patient clinical data management system based on feature extraction, including: A data acquisition module 210 is used to acquire clinical data of a diabetic patient, where the clinical data includes blood sugar data and behavior data of the patient; A first feature extraction module 220 is used to extract behavior matching features between different behaviors of diabetic patients through patient behavior data; The second feature extraction module 230 is used to extract the correlation features between different behaviors of diabetic patients and blood sugar fluctuations through the patient blood sugar data and patient behavior data; The risk warning module 240 is used to evaluate the blood sugar fluctuation risk according to the behavior matching characteristics and associated characteristics of the diabetic patient and to issue a blood sugar fluctuation risk warning according to the blood sugar fluctuation risk assessment result.

[0025] In the embodiment of the present application, the first feature extraction module 220 is used to extract the behavior matching features between different behaviors of diabetic patients through the patient behavior data. The specific steps of extracting the behavior matching features between different behaviors of diabetic patients include: Acquire patient behavior data, which includes dining behavior data, medication behavior data, and exercise behavior data of diabetic patients; Extracting behavior matching features between different behaviors through patient behavior data, the behavior matching features include first behavior matching features, second behavior matching features and third behavior matching features; Among them, the first behavior matching feature is used to evaluate the matching degree between the dining behavior and medication behavior of diabetic patients, and the first behavior matching feature is calculated through the dining behavior data and medication behavior data; The second behavior matching feature is used to evaluate the matching degree between the dining behavior and the exercise behavior of the diabetic patients. The second behavior matching feature is calculated through the dining behavior data and the exercise behavior data. The third behavior matching feature is used to evaluate the matching degree between the medication behavior and exercise behavior of diabetic patients. The third behavior matching feature is calculated through the medication behavior data and exercise behavior data. The specific steps of calculating the first behavior matching feature through the meal behavior data and the medication behavior data include: Obtaining the patient's dining behavior data and medication behavior data from the patient's behavioral data, wherein the dining behavior data includes dining time, meal weight, and meal food type, and the medication behavior data includes medication time, medication dosage, and medication type; The difference between the meal time and the medication time is taken as the medication time interval, and the medication time matching degree is obtained by comparing the medication time interval with the optimal medication time interval corresponding to the drug type. When the medication time interval is less than or equal to the optimal medication time interval, the medication time matching degree is 1, and when the medication time interval is greater than the optimal medication time interval, the medication time matching degree is 0; The carbohydrate intake is determined by the weight and type of meal food, and the dosage matching degree is calculated by the carbohydrate intake content and the standard carbohydrate content corresponding to the dosage. The dosage matching degree calculation formula is as follows: ; In the formula Indicates the matching degree of drug dosage. Indicates carbohydrate intake. Indicates the standard carbohydrate content corresponding to the dosage; The mean of medication time matching and medication dosage matching is taken as the first row matching feature; The specific steps of calculating the second behavior matching feature through the dining behavior data and the exercise behavior data include: Acquire dining behavior data and exercise behavior data, wherein the dining behavior data includes dining time, meal weight and meal type, and the exercise behavior data includes exercise start time and exercise time set; The difference between the meal time and the exercise start time is taken as the exercise time interval, and the second behavior matching feature is obtained by comparing the exercise time interval with the recommended exercise time interval. When the exercise time interval is greater than the recommended exercise time interval, the second behavior matching feature is 1, and when the exercise time interval is less than or equal to the exercise time interval, the second behavior matching feature is 0; The specific steps of calculating the third behavior matching feature through medication behavior data and exercise behavior data include: Obtain medication behavior data and exercise behavior data, the medication behavior data includes medication time, medication dosage and medication type, and the exercise behavior data includes exercise start time and exercise time set; Obtain the maximum drug effect time set by the maximum drug effect time corresponding to the medication time and drug type, and calculate the number of elements of the intersection of the movement time set and the maximum drug effect time set and the number of elements of the union of the movement time set and the maximum drug effect time set; Calculate the matching features of the third line by the number of intersection elements and the number of union elements , ,in, Represents the number of elements in the intersection of the exercise time set and the maximum drug effect time set, Represents the number of elements in the union of the exercise time set and the maximum drug effect time set, Indicates that the third line is a matching feature.

[0026] In the embodiment of the present application, the second feature extraction module 230 is used to extract the correlation features between different behaviors of diabetic patients and blood sugar fluctuations through the patient blood sugar data and the patient behavior data. The specific steps of extracting the correlation features between different behaviors of diabetic patients and blood sugar fluctuations include: Obtaining patient blood sugar data and patient behavior data, wherein the patient behavior data includes the dining behavior data, medication behavior data and exercise behavior data of diabetic patients; Extracting behavioral time points through patient behavioral data, the behavioral time points include the patient's meal behavior time points, medication behavior time points, and exercise behavior time points; The blood sugar fluctuation time points are extracted from the patient's blood sugar data and the correlation characteristics between each behavior type and blood sugar fluctuation are calculated in combination with the patient's behavior time points. The correlation characteristic calculation formula is as follows: ; In the formula Indicates patient behavior No. Behavior time point, Indicates the patient behavior type, represent eating behavior, medication behavior, and exercise behavior, respectively. Indicates patient behavior The number of behavioral time points, Indicates Blood sugar fluctuation time point, Indicates the number of blood sugar fluctuation time points, Indicates patient behavior No. The time difference between a behavior time point and the most recent blood sugar fluctuation after the behavior time point, Indicates the data monitoring cycle, Indicates patient behavior Characteristics associated with blood sugar fluctuations.

[0027] In the embodiment of the present application, the risk warning module 240 is used to evaluate the blood sugar fluctuation risk according to the behavior matching characteristics and the associated characteristics of the diabetic patient and to issue a blood sugar fluctuation risk warning according to the blood sugar fluctuation risk evaluation result. The specific steps of evaluating the blood sugar fluctuation risk and issuing a blood sugar fluctuation risk warning according to the blood sugar fluctuation risk evaluation result include: Obtaining behavioral matching features and correlation features between different behaviors and blood sugar fluctuations; The blood sugar fluctuation risk index is calculated through the behavior matching characteristics and the association characteristics. The blood sugar fluctuation risk index is used to assess the patient's blood sugar fluctuation risk. The blood sugar fluctuation risk index calculation formula is as follows: ; In the formula Indicates that the first line is a matching feature. Indicates that the second line is a matching feature. Indicates that the third line is a matching feature. Indicates the correlation characteristics between patients' meal behavior and blood sugar fluctuations, Indicates the correlation characteristics between patients' medication behavior and blood sugar fluctuations, Indicates the correlation characteristics between patients' exercise behavior and blood sugar fluctuations, represents the blood sugar fluctuation risk index; When the blood sugar fluctuation risk index is greater than the preset blood sugar fluctuation risk threshold, a blood sugar fluctuation risk warning is issued. When the blood sugar fluctuation risk index is less than or equal to the preset blood sugar fluctuation risk threshold, no blood sugar fluctuation risk warning is issued.

[0028] The above-mentioned parameters and steps for each unit module to implement corresponding functions in the feature extraction-based clinical data management system for diabetic patients of the present application can refer to the parameters and steps in the embodiment of the feature extraction-based clinical data management method for diabetic patients mentioned above, and will not be repeated here.

[0029] like Figure 4As shown, an embodiment of the present invention further provides an electronic device 300, including a memory 320 for storing a computer program 322; and a processor 310 for executing the computer program 322 to implement a clinical data management method for diabetic patients based on feature extraction as in any of the above embodiments.

[0030] It should be noted that Figure 4 It is a structural diagram of an electronic device 300 according to an exemplary embodiment. The content in the diagram cannot be regarded as any limitation on the scope of application of the present invention.

[0031] Specifically, the electronic device 300 may include: at least one processor 310, at least one memory 320, a power supply 330, a communication interface 340, an input / output interface 350, and a communication bus 360. The memory 320 is used to store a computer program 322, which is loaded and executed by the processor 310 to implement the relevant steps in the method for managing clinical data of diabetic patients based on feature extraction disclosed in any of the aforementioned embodiments. In addition, the electronic device 300 in the embodiment of the present invention may be specifically an electronic computer.

[0032] In an embodiment of the present invention, a power supply 330 is used to provide a working voltage for each hardware device on the electronic device 300; the communication interface 340 can create a data transmission channel between the electronic device 300 and an external device, and the communication protocol it follows is any communication protocol that can be applied to the technical solution of the present invention, and is not specifically limited here; the input and output interface 350 is used to obtain external input data or output data to the outside world, and its specific interface type can be selected according to specific application needs and is not specifically limited here.

[0033] In addition, the memory 320, as a carrier for storing resources, can be a read-only memory, a random access memory, a disk or an optical disk, etc. The resources stored thereon can include an operating system 321, a computer program 322, etc., and the storage method can be temporary storage or permanent storage.

[0034] The operating system 321 is used to manage and control the hardware devices and computer programs on the electronic device 300, and can be Windows Server, Netware, Unix, Linux, etc. In addition to the computer program 322 that can be used to complete the feature extraction-based clinical data management method for diabetic patients performed by the electronic device 300 disclosed in any of the aforementioned embodiments, the computer program 322 can further include a computer program 322 that can be used to complete other specific tasks.

[0035] An embodiment of the present invention further provides a computer-readable storage medium for storing a computer program 322 , wherein the computer program 322 , when executed by the processor 310 , implements a method for managing clinical data of diabetic patients based on feature extraction as described in any of the above embodiments. For example, the computer readable storage medium can be a read-only memory (ROM), a random access memory (RAM), a compact disc (CD-ROM), a magnetic tape, a floppy disk, an optical data storage device, etc.

Claims

1. A method for managing clinical data of diabetic patients based on feature extraction, characterized in that: The steps include: Acquire clinical data of diabetic patients, where the clinical data includes blood sugar data and behavior data of the patients; Extract behavioral matching features between different behaviors of diabetic patients through patient behavior data; Extract the correlation characteristics between different behaviors of diabetic patients and blood sugar fluctuations through patient blood sugar data and patient behavior data; The blood sugar fluctuation risk is assessed based on the behavioral matching characteristics and associated characteristics of diabetic patients, and blood sugar fluctuation risk warnings are issued based on the blood sugar fluctuation risk assessment results.

2. The method for managing clinical data of diabetic patients based on feature extraction according to claim 1, characterized in that: The specific steps of extracting the behavior matching features between different behaviors of diabetic patients include: Acquire patient behavior data, which includes dining behavior data, medication behavior data, and exercise behavior data of diabetic patients; Extracting behavior matching features between different behaviors through patient behavior data, the behavior matching features include first behavior matching features, second behavior matching features and third behavior matching features; Among them, the first behavior matching feature is used to evaluate the matching degree between the dining behavior and medication behavior of diabetic patients, and the first behavior matching feature is calculated through the dining behavior data and medication behavior data; The second behavior matching feature is used to evaluate the matching degree between the dining behavior and the exercise behavior of the diabetic patients. The second behavior matching feature is calculated through the dining behavior data and the exercise behavior data. The third behavior matching feature is used to evaluate the matching degree between the medication behavior and exercise behavior of diabetic patients. The third behavior matching feature is calculated through the medication behavior data and the exercise behavior data.

3. The method for managing clinical data of diabetic patients based on feature extraction according to claim 2, characterized in that: The specific steps of calculating the first behavior matching feature through the dining behavior data and the medication behavior data include: Obtaining the patient's dining behavior data and medication behavior data from the patient's behavioral data, wherein the dining behavior data includes dining time, meal weight, and meal food type, and the medication behavior data includes medication time, medication dosage, and medication type; The difference between the meal time and the medication time is taken as the medication time interval, and the medication time matching degree is obtained by comparing the medication time interval with the optimal medication time interval corresponding to the drug type. When the medication time interval is less than or equal to the optimal medication time interval, the medication time matching degree is 1, and when the medication time interval is greater than the optimal medication time interval, the medication time matching degree is 0; The carbohydrate intake is determined by the weight and type of meal food, and the dosage matching degree is calculated by the carbohydrate intake content and the standard carbohydrate content corresponding to the dosage. The dosage matching degree calculation formula is as follows: ； In the formula Indicates the matching degree of drug dosage. Indicates carbohydrate intake. Indicates the standard carbohydrate content corresponding to the dosage; The mean of the medication time matching degree and the medication dosage matching degree is taken as the first row matching feature.

4. The method for managing clinical data of diabetic patients based on feature extraction according to claim 2, characterized in that: The specific steps of calculating the second behavior matching feature through the dining behavior data and the exercise behavior data include: Acquire dining behavior data and exercise behavior data, wherein the dining behavior data includes dining time, meal weight and meal type, and the exercise behavior data includes exercise start time and exercise time set; The difference between the meal time and the exercise start time is taken as the exercise time interval, and the second behavior matching feature is obtained by comparing the exercise time interval with the recommended exercise time interval. When the exercise time interval is greater than the recommended exercise time interval, the second behavior matching feature is 1, and when the exercise time interval is less than or equal to the exercise time interval, the second behavior matching feature is 0.

5. The method for managing clinical data of diabetic patients based on feature extraction according to claim 2, characterized in that: The specific steps of calculating the third behavior matching feature through the medication behavior data and the exercise behavior data include: Obtain medication behavior data and exercise behavior data, the medication behavior data includes medication time, medication dosage and medication type, and the exercise behavior data includes exercise start time and exercise time set; Obtain the maximum drug effect time set by the maximum drug effect time corresponding to the medication time and drug type, calculate the number of elements of the intersection of the movement time set and the maximum drug effect time set, and the number of elements of the union of the movement time set and the maximum drug effect time set; Calculate the matching features of the third line by the number of intersection elements and the number of union elements , ,in, Represents the number of elements in the intersection of the exercise time set and the maximum drug effect time set, Represents the number of elements in the union of the exercise time set and the maximum drug effect time set, Indicates that the third line is a matching feature.

6. The method for managing clinical data of diabetic patients based on feature extraction according to claim 1, characterized in that: The specific steps of extracting the correlation characteristics between different behaviors of diabetic patients and blood sugar fluctuations include: Obtaining patient blood sugar data and patient behavior data, wherein the patient behavior data includes the dining behavior data, medication behavior data and exercise behavior data of diabetic patients; Extracting behavioral time points through patient behavioral data, the behavioral time points include the patient's meal behavior time points, medication behavior time points, and exercise behavior time points; The blood sugar fluctuation time points are extracted from the patient's blood sugar data and the correlation characteristics between each behavior type and blood sugar fluctuation are calculated in combination with the patient's behavior time points. The correlation characteristic calculation formula is as follows: ； In the formula Indicates patient behavior No. Behavior time point, Indicates the patient behavior type, represent eating behavior, medication behavior, and exercise behavior, respectively. Indicates patient behavior The number of behavioral time points, Indicates Blood sugar fluctuation time point, Indicates the number of blood sugar fluctuation time points, Indicates patient behavior No. The time difference between a behavior time point and the most recent blood sugar fluctuation after the behavior time point, Indicates the data monitoring cycle. Indicates patient behavior Characteristics associated with blood sugar fluctuations.

7. The method for managing clinical data of diabetic patients based on feature extraction according to claim 1, characterized in that: The specific steps of assessing the blood sugar fluctuation risk and providing a blood sugar fluctuation risk warning according to the blood sugar fluctuation risk assessment result include: Obtaining behavioral matching features and correlation features between different behaviors and blood sugar fluctuations; The blood sugar fluctuation risk index is calculated through the behavior matching characteristics and the association characteristics. The blood sugar fluctuation risk index is used to assess the patient's blood sugar fluctuation risk. The blood sugar fluctuation risk index calculation formula is as follows: ； In the formula Indicates that the first line is a matching feature. Indicates that the second line is a matching feature. Indicates that the third line is a matching feature. Indicates the correlation characteristics between patients' meal behavior and blood sugar fluctuations, Indicates the correlation characteristics between patients' medication behavior and blood sugar fluctuations, Indicates the correlation characteristics between patients' exercise behavior and blood sugar fluctuations, represents the blood sugar fluctuation risk index; When the blood sugar fluctuation risk index is greater than the preset blood sugar fluctuation risk threshold, a blood sugar fluctuation risk warning is issued. When the blood sugar fluctuation risk index is less than or equal to the preset blood sugar fluctuation risk threshold, no blood sugar fluctuation risk warning is issued.

8. A diabetes patient clinical data management system based on feature extraction, applied to the diabetes patient clinical data management method based on feature extraction according to any one of claims 1 to 7, characterized in that: The system comprises: A data acquisition module is used to acquire clinical data of diabetic patients, where the clinical data includes blood sugar data and behavior data of patients; A first feature extraction module is used to extract behavior matching features between different behaviors of diabetic patients through patient behavior data; The second feature extraction module is used to extract the correlation features between different behaviors of diabetic patients and blood sugar fluctuations through the patient blood sugar data and patient behavior data; The risk warning module is used to evaluate the blood sugar fluctuation risk based on the behavioral matching characteristics and associated characteristics of diabetic patients and to issue blood sugar fluctuation risk warning based on the blood sugar fluctuation risk assessment results.

9. An electronic device, comprising: A processor and a memory, wherein the memory stores a computer program that can be called by the processor; characterized in that the processor executes the clinical data management method for diabetic patients based on feature extraction as described in any one of claims 1-7 by calling the computer program stored in the memory.

10. A computer-readable storage medium, characterized in that: Instructions are stored, and when the instructions are executed on a computer, the computer is caused to execute the method for managing clinical data of diabetic patients based on feature extraction as described in any one of claims 1 to 7.

Images