A method for predicting blood sugar changes in patients with type 2 diabetes using mathematical models

Abstract

The present invention discloses a method for predicting blood glucose changes in type II diabetic patients through a mathematical model; in the initial stage of model operation, the blood glucose prediction value is automatically calculated according to the recorded eating situation of the patient, and is matched with the self-measured blood glucose value after a meal input by the user. For the first match, corresponding values are generated according to the results, thereby forming two sets of mathematical models for the next match. If the blood glucose value calculated according to a certain corresponding value can match the self-measured blood glucose value of subsequent users, then this value is retained; if it cannot match, the two self-measured blood glucose values are combined and calculated again until the match is successful. Through the interaction of insulin, glucagon, blood glucose, and glycogen, a mathematical model is constructed, and exogenous substances such as food, drugs, and insulin are added to the mathematical model, so as to predict the postprandial blood glucose by recording the food eaten by the patient every day. At the same time, the patient can calculate the blood glucose change after medication through the selection of the doses of relevant drugs and insulin in the model.

Description

Technical Field

[0001] The present invention relates to the technical field related to information technology, and specifically relates to a method for implementing a mathematical model for predicting blood glucose changes in type II diabetic patients through an automatic learning method. Background Art

[0002] Diabetes is a group of metabolic disorders of carbohydrates, proteins, and fats caused by absolute or relative insulin deficiency and insulin utilization disorders. In 2021, the total number of diabetic patients in China reached 129.8 million. The research results show that diabetes is an important health problem in China. Continuous monitoring and effective control are required to reduce its burden. At present, the vast majority of blood glucose measurements for diabetic patients use fingertip blood measurement. Severe patients measure blood glucose 3-4 times a day, causing great pain to the patients. At the same time, some patients, due to low educational level, less medical knowledge, and poor compliance, cannot well control blood glucose changes through diet and medication, resulting in fluctuating conditions and affecting the treatment effect. Summary of the Invention

[0003] Therefore, to solve the above deficiencies, the present invention provides a method for implementing a mathematical model for predicting blood glucose changes in type II diabetic patients through an automatic learning method; the present invention constructs a mathematical model through the interaction of insulin, glucagon, blood glucose, and glycogen, and adds exogenous substances such as food, drugs, and insulin to the mathematical model, so as to predict postprandial blood glucose by recording the food eaten by the patient every day. At the same time, the patient can calculate the blood glucose change after taking the medicine through the selection of the doses of relevant drugs and insulin in the model. And an automatic learning and adjustment function is added to the system. In the first few times of use, according to the blood glucose changes after the patient eats and takes the medicine, the relevant parameters in the mathematical model are automatically adjusted, so that the blood glucose changes generated by the model in the later stage are more consistent with the patient's blood glucose changes, thereby reducing the number of times the patient measures blood glucose and alleviating the patient's pain.

[0004] The present invention is implemented as follows. A method for implementing the prediction of blood glucose changes in type II diabetic patients through a mathematical model is characterized in that the specific implementation process is as follows;

[0005] Step 1, when used for the first time, all variable parameters in the system, including the postprandial blood glucose intake rate G1, the exogenous insulin input rate G4, the constant C1, the constant C4, and the insulin hypoglycemic constant a 14 are all default values, and this value is derived from the adult average values recorded in multiple documents;

[0006] Step 2, the patient measures the blood glucose value before breakfast and inputs it into the system. This blood glucose is the initial value x of x1 10 for calculating subsequent blood glucose values;

[0007] Step 3: Subsequently, the patient has breakfast and enters the type and weight of the breakfast into the system. The system calculates the post-meal blood glucose intake rate G1 and substitutes it into the mathematical model.

[0008] Step 4: The patient measures the blood glucose x 11 two hours after the meal and enters it into the system. The system compares the value of x 11 with the x1 value calculated by the system.

[0009] Step 5: If the difference between the value of x 11 and the blood glucose concentration x1 value is too large, the system automatically adjusts the two variable parameters of the insulin hypoglycemic constant a 14 and the constant C4. On the premise of ensuring that the x1 value = x 11 value, at least 10 sets of a 14 and C4 values are generated for subsequent mathematical model calculations.

[0010] Step 6: The system calculates according to the 10 sets of a 14 and C4 values calculated in Step 5 and predicts the blood glucose value before lunch for the patient respectively.

[0011] Step 7: The patient self-measures the blood glucose value x 12 before lunch and records it. The system compares x 12 with the 10 sets of x1 values, and retains the two sets of a 12 and C4 parameters that are closest to x 14 .

[0012] Step 8: On the premise of ensuring that the x1 value = x 12 value, the system automatically adjusts the two variable parameters of the insulin hypoglycemic constant a 14 and the constant C4, and each generates at least 10 sets of a 14 and C4 values for subsequent mathematical model calculations.

[0013] Step 9: The patient has lunch, enters the type and weight of the lunch into the mathematical model, and measures the blood glucose 2 hours after the meal and enters it into the system.

[0014] Step 10: Repeat Steps 7 and 8 until the blood glucose error before and after meals does not exceed ±0.1 mmol / L for five consecutive times.

[0015] Step 11: At this time, the insulin hypoglycemic constant a 14 and the C4 value are basically the same as the actual values of the patient, and the system will continue to calculate the subsequent blood glucose changes according to this value.

[0016] Step 12: When the patient completes the measurement of the insulin hypoglycemic constant a 14 and the constant C4, the system adjusts the measured drug tolerance of the patient.

[0017] Step 13: Measure the blood glucose before the patient takes the medicine and input it into the system. Then, take the medicine according to the doctor's advice and record the type and dosage of the medicine. After that, measure the blood glucose and record x. 15 ;

[0018] Step 14: According to the different drug action mechanisms, the system adjusts the constant C1, the constant C4, and the insulin hypoglycemic constant a respectively with preset values. 14、 For the exogenous insulin input rate G4, calculate the value of x1.

[0019] Step 15: Compare the measured value of x 15 with the value of x1 calculated by the system. According to the type of medicine, adjust the body's tolerance to the medicine, the postprandial blood glucose intake rate C1, the exogenous insulin input rate C4, and the insulin hypoglycemic constant a 14 and the constant G4. The adjustment method is similar to Step 5.

[0020] Step 16: Repeat Steps 14 - 16 until the difference between the measured value of blood glucose after taking the medicine and the calculated value by the system does not exceed 0.1 mmol / L; (According to the formula structure, it is expected that a total of 3 - 5 blood glucose measurements are required to complete);

[0021] Step 17: At this time, the body's tolerance to the medicine is basically the same as that of the patient. The system will calculate the subsequent blood glucose changes according to this value.

[0022] Step 18: If the difference between the measured value of blood glucose in three consecutive times and the calculated value exceeds 0.5 mmol / L, consider the change of the patient's condition, and the insulin hypoglycemic constant a 14 and the value of the constant C4 change, then repeat Steps 1 - 17 for recalculation.

[0023] The present invention has the following advantages: The present invention provides a method for implementing a mathematical model for predicting blood glucose changes in type II diabetic patients through an automatic learning method; by the interaction of insulin, glucagon, blood glucose, and glycogen, the present invention constructs a mathematical model and adds exogenous substances such as food, medicine, and insulin into the mathematical model, so as to predict the postprandial blood glucose by recording the food eaten by the patient every day. At the same time, the patient can calculate the blood glucose change after taking the medicine through the selection of the dosages of relevant medicine and insulin in the model. And an automatic learning and adjustment function is added to the system. In the first few times of use, according to the blood glucose changes after the patient eats and takes the medicine, the relevant parameters in the mathematical model are automatically adjusted, so that the blood glucose changes generated by the model in the later stage are more in line with the patient's blood glucose changes, thereby reducing the number of times the patient measures blood glucose and alleviating the patient's pain. Brief Description of the Drawings

[0024] Figure 1 is a schematic diagram of the process of the mathematical model of the present invention;

[0025] Figure 2Schematic diagram of the glycemic index of food and the corresponding G1 value;

[0026] Figure 3 Effect of the drug on a 14 and c1;

[0027] Figure 4 Schematic diagram of the effect of insulin on G4 and its duration. Specific implementation manner

[0028] The following will combine the attached Figures 1 - 4 This invention will be described in detail below. The technical solutions in the embodiments of the invention will be clearly and completely described. Obviously, the described embodiments are only a part of the embodiments of the invention, rather than all of the embodiments. Based on the embodiments of the invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the invention.

[0029] This invention provides a method for implementing a mathematical model for predicting blood glucose changes in type II diabetic patients through an automatic learning method, which is implemented as follows;

[0030] According to the definition of diabetes, type II diabetes is caused by insulin resistance, insulin secretion, and / or insulin utilization disorders. The two parameters related to this in the mathematical model are a 14 , C4. The former is the insulin resistance and utilization parameter, and the latter is the insulin synthesis and secretion parameter. There are a total of 5 parameters related to the food eaten by the patient, the drugs taken, and the insulin injected, which are: G1, G4, C1, C4, a 14 , where G1 is the food eaten, G4 is the insulin injected, and the remaining 3 are the parameters related to the drugs taken.

[0031] At the initial stage of model operation, the system automatically calculates the blood glucose prediction value according to the recorded eating situation of the patient and matches it with the self-measured blood glucose value input by the user. According to the result of the first match, the corresponding values of a 14 , C4 are generated, thus forming two sets of mathematical models for the next match. If the blood glucose value calculated according to a certain corresponding value of a 14 , C4 can match the self-measured blood glucose value of the subsequent user, then retain this value; if it cannot match, combine the two self-measured blood glucose values and calculate again until the match is successful. The matching process is shown in Figure 1 .

[0032] For each food, calculate the corresponding G1 parameter according to the glycemic index. The glycemic index is shown in Figure 2 . The operator selects the corresponding food and consumption amount in the table according to the eating situation. The system calculates the G1 value according to the food type and consumption amount, substitutes it into the model for calculation, and then generates the corresponding blood glucose prediction value.

[0033] According to their different pharmacological effects, each drug can act on different parameters in the model. For the specific affected parameters, see Figure 3 . After taking the drug, the system adjusts the corresponding parameters according to the pharmacokinetic characteristics of the drug, inputs them into the model, and calculates the corresponding blood glucose prediction value.

[0034] Insulin is divided into short-acting, intermediate-acting and long-acting insulin. Each type of insulin has a different duration and different effects on G4. For the specific effects, see Figure 4 . After the patient injects insulin, the system automatically calculates the corresponding blood glucose prediction value according to the duration of insulin and its effect on G4.

[0035] The following is an explanation of the specific implementation process of the present invention;

[0036] Step 1, when first used, all variable parameters in the system, including G1, G4, C1, C4, a 14 are default values, which are derived from the adult average values recorded in multiple literatures;

[0037] Step 2, the patient measures the blood glucose value before breakfast in the morning and inputs it into the system. This blood glucose is the initial value x of x1 10 , which is used to calculate subsequent blood glucose values;

[0038] Step 3, then the patient has breakfast and inputs the type and weight of the breakfast into the system. The system calculates and generates G1 and substitutes it into the mathematical model;

[0039] Step 4, the patient measures the blood glucose x 11 two hours after the meal and inputs it into the system. The system compares the x 11 value with the x1 value calculated by the system;

[0040] Step 5, if the x 11 value and the x1 value differ too much, the system automatically adjusts the two variable parameters of a 14 and C4. On the premise of ensuring that the x1 value = x 11 value, at least 10 sets of a 14 and C4 values are generated for subsequent mathematical model calculations;

[0041] Step 6, the system calculates according to the 10 sets of a 14 and C4 values calculated in Step 5, and respectively predicts the blood glucose value before lunch of the patient;

[0042] Step 7, the patient self-measures the blood glucose value x 12 before lunch and records it. The system compares x 12 with the 10 sets of x1 values, and retains the two sets of a 14 and C4 parameters that are closest to x12;

[0043] Step 8, on the premise of ensuring that the value of x1 = x 12 value, the system automatically adjusts the two variable parameters of a 14 and C4, and generates at least 10 groups of a 14 and C4 values for subsequent mathematical model calculations;

[0044] Step 9, the patient takes lunch, enters the type and weight of lunch into the mathematical model, and measures the blood glucose 2 hours after the meal and enters it into the system;

[0045] Step 10, repeat steps 7 and 8 until the blood glucose error before and after meals does not exceed ±0.1 mmol / L for five consecutive times (it is estimated that a total of 5 - 10 blood glucose measurements are required to complete according to the structure of Formula 1 - 4);

[0046] Step 11, at this time, the values of a 14 and C4 are basically the same as the actual values of the patient, and the system will continue to calculate the subsequent blood glucose changes according to these values;

[0047] Step 12, when the patient completes the measurement of a 14 and C4, the system will adjust the drug tolerance of the patient;

[0048] Step 13, measure the blood glucose before the patient takes the medicine and enter it into the system, then take the medicine according to the doctor's advice and record the type and dosage of the medicine, and then measure the blood glucose and record x 15 ;

[0049] Step 14, according to the different drug action mechanisms, the system adjusts C1, C4, a 14、 G4 (see Table 2 and Table 3) respectively with preset values, and calculates the x1 value (Formula 1 involved in the invention);

[0050] Step 15, compare the measured x15 value with the x1 value calculated by the system, and adjust the drug tolerance C1, C4, a 14 and G4 of the body according to the type of medicine, and the adjustment method is similar to step 5;

[0051] Step 16, repeat steps 14 - 16 until the difference between the measured blood glucose value after taking the medicine and the calculated value by the system does not exceed 0.1 mmol / L (it is estimated that a total of 3 - 5 blood glucose measurements are required to complete according to the structure of Formula 1 - 4);

[0052] Step 17, at this time, the drug tolerance of the body is basically the same as the actual situation of the patient, and the system will calculate the subsequent blood glucose changes according to this value;

[0053] Step 18, if the difference between the measured blood glucose value and the calculated value exceeds 0.5 mmol / L for three consecutive times, consider the change of the patient's condition, a 14If the C4 value changes, repeat steps 1-17 for recalculation.

[0054] The formula involved in the present invention is:

[0055]

[0056] The meanings of the characters in the model are shown in the following table

[0057]

[0058] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for predicting blood glucose changes in type II diabetic patients through a mathematical model, characterized in that; The specific implementation process is as follows; Step 1. When used for the first time, all variable parameters in the system, including the postprandial blood glucose intake rate G1, the exogenous insulin input rate G4, the constant C1, the constant C4, and the insulin hypoglycemic constant a 14 are all default values, which are derived from the average values of adults recorded in multiple documents; Step 2, the patient measures the blood glucose value before breakfast in the morning and inputs it into the system. This blood glucose value is the initial value x of x1, which is used to calculate subsequent blood glucose values; 10 ​ Step 3: Subsequently, the patient has breakfast and enters the type and weight of the breakfast into the system. The system calculates the postprandial blood glucose intake rate G1 and substitutes it into the mathematical model; Step 4, the patient measures the blood glucose x two hours after a meal 11 , and inputs it into the system. The system will compare the value of x 11 with the value of x1 calculated by the system; Step 5, if the difference between the x 11 value and the blood glucose concentration x1 value is too large, the system automatically adjusts the insulin hypoglycemic constant a 14 , the two variable parameters of the constant C4, and generates at least 10 groups of a 11 and C4 values on the premise of ensuring that the x1 value = x 14 value for subsequent mathematical model calculations; Step 6: The system calculates according to the 10 groups of a 14 and C4 values calculated in Step 5 to respectively predict the pre-lunch blood glucose values of the patients; Step 7, the patient self-tests the blood glucose value x before lunch 12 and records it, and the system compares x 12 with 10 groups of x1 values, and retains the two groups a 12 that are closest to x 14 and the C4 parameter; Step 8, on the premise of ensuring that the value of x1 = x 12 value, the system automatically adjusts the two variable parameters of the insulin hypoglycemic constant a 14 , constant C4, and generates at least 10 sets of a 14 and C4 values for subsequent mathematical model calculations; Step 9: The patient has lunch, enters the type and weight of the lunch into the mathematical model, and measures the blood glucose 2 hours after the meal and enters it into the system; Step 10: Repeat steps 7 and 8 until the pre- and post-meal blood glucose errors do not exceed ±0.1 mmol / L for five consecutive times; Step 11, at this time, the insulin hypoglycemic constant a 14 is basically the same as the value of constant C4 and the actual value of the patient, and the system will continue to calculate the subsequent blood glucose changes according to this value; Step 12, when the patient completes the measurement of the insulin hypoglycemic constant a 14 and the constant C4, the system will adjust by measuring the patient's tolerance to the drug; Step 13: Measure the blood glucose of the patient before taking the medicine and input it into the system. Then, take the medicine according to the doctor's advice and record the type and dosage of the medicine. Then, measure the blood glucose and record x 15 ; Step 14: The system adjusts the constant C1, the constant C4, and the insulin hypoglycemic constant a respectively with preset values according to different drug action mechanisms. 14、 For the exogenous insulin infusion rate G4, calculate the value of x1. Step 15, measure x 15 Compare the measured value of x with the value of x1 calculated by the system, and adjust the drug tolerance of the body, the postprandial blood glucose intake rate C1, the exogenous insulin infusion rate C4, and the insulin hypoglycemic constant a 14 , and the constant G4 according to the drug type. The adjustment method is carried out in the same way as in Step 5; Step 16: Repeat steps 14 - 16 until the difference between the measured blood glucose value after taking the medicine and the calculated value of the system does not exceed 0.1 mmol / L; Step 17: At this time, the drug tolerance of the body is basically the same as that of the patient, and the system will calculate the subsequent blood glucose changes according to this value; Step 18, if the difference between the measured blood glucose value and the calculated value exceeds 0.5 mmol / L for three consecutive times, consider the change in the patient's condition and the change in the insulin hypoglycemic constant a 14 and the value of constant C4, then repeat steps 1-17 for recalculation; The formulas involved in the above mathematical model are: （1） （2） （3） （4） Among them: blood glucose concentration x1, glycogen concentration x2, glucagon concentration x3, insulin concentration x4, postprandial blood glucose intake rate G1, postprandial glycogen intake rate G2, exogenous intake rate G3, exogenous insulin input rate G4, Glucagon blood sugar elevation constant , in vivo glycogen synthesis constant , time constant t, constant C1, constant C2, constant C3, constant C4, Constant ,Constant ,Constant ,Constant , Insulin hypoglycemic constant a 14 , Hepatic glycogenolysis constant a 23 , Glucagon loss rate a during glycogenolysis 32 , Insulin loss rate a during glycogenesis 41 , glycolysis constant k1, glycogen degradation constant k2, glucagon enzyme degradation k3, insulin index exhaustion k4, Other fixed consumption k 01 , other fixed consumption k 02 , other fixed consumption k 03 , other fixed consumption k 04 .

Images