An automatic detection system for symptoms of diabetes

Abstract

The present application relates to the technical field of physiological measurement, in particular to a kind of diabetes symptom automatic detection system, system includes: blood glucose fluctuation analysis module, polydipsia and polyuria symptom detection module, recession level evaluation module, progress stage identification module.In the present application, through the correlation analysis of real-time blood glucose monitoring and patient diet state, the change of blood glucose level is accurately captured, the development trend of disease is predicted, the real-time of disease monitoring is enhanced, the medical intervention is more timely and effective, through the continuous monitoring of the amount of water and urine of patient, the symptoms of polydipsia and polyuria of diabetic patients are found in time, and the correlation between target symptom and diabetes progression is determined through data analysis, the identification ability of early diabetic complications is enhanced, combined with the evaluation of organ function level, the signs of organ damage are identified, help to adjust treatment strategy, prevent the development of complications, improve the real-time and accuracy of disease monitoring, reduce the complexity and cost of long-term treatment.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of physiological measurement, in particular to an automatic detection system for diabetic symptoms. BACKGROUND

[0002] The technical field of physiological measurement focuses on detecting, recording and interpreting various physiological signals of the human body, which is applied to monitoring heart activity, blood pressure, blood glucose, blood oxygen saturation, body temperature and various important physiological parameters, through the use of sensors, wearable devices and non-invasive monitoring systems, to help medical professionals obtain the health status of patients in real time, and assist in diagnosis, monitoring of disease progression, monitoring of treatment effect and prevention of development of chronic diseases, combined with data processing and analysis technology, to extract various useful information from complex data, to realize long-term health management and acute condition monitoring.

[0003] Among them, an automatic detection system for diabetic symptoms focuses on automatically monitoring and identifying various related symptoms of diabetes, including real-time tracking of patients' blood glucose levels, physical reactions and various physiological indicators indicating changes in the disease, to help understand the blood glucose status of patients in real time, to realize automatic data collection and analysis through the integration of sensing technology and intelligent algorithms, to provide key information to doctors and patients in a timely manner and assist in medical decision-making, to help manage diabetes and prevent complications, to realize early intervention and reduce health risks, and to improve the treatment effect and quality of life of patients.

[0004] The traditional automatic detection system for diabetic symptoms has steps in the management and prevention of diabetes, focusing on the monitoring of a single indicator or providing data at a specific time point, lacking in-depth analysis of the correlation between data and the ability to predict long-term trends, and being unable to analyze the trend of blood glucose changes and the interaction with various physiological parameters, including changes in diet and urine volume, resulting in medical professionals being unable to make the best medical decisions in the absence of real-time access to comprehensive information, relying on regular medical examinations in monitoring organ decline, delaying the diagnosis of early organ damage and missing the best treatment opportunity, and lacking in disease progression prevention measures. SUMMARY

[0005] The purpose of the present application is to solve the shortcomings in the prior art, and an automatic detection system for diabetic symptoms is proposed.

[0006] In order to achieve the above-mentioned purpose, the present application adopts the following technical scheme: an automatic detection system for diabetic symptoms comprises:

[0007] The blood glucose fluctuation analysis module analyzes the blood glucose fluctuation pattern of the patient based on real-time blood glucose monitoring data, analyzes the influence of nutrient intake on blood glucose fluctuation in combination with the diet status of the patient, predicts the development trend of the blood glucose level of the patient, and generates a blood glucose feature data set;

[0008] The polydipsia and polyuria symptom detection module is based on the blood glucose feature data set, collects the water intake and urine volume data of the patient, evaluates the polyuria symptom level of the patient, calculates the correlation between the polyuria symptom feature and the progression stage of diabetes according to the correlation between the urine volume level and the blood glucose feature, and outputs the polyuria symptom analysis result;

[0009] The decline level evaluation module utilizes the polyuria symptom analysis result, collects a plurality of physiological feature data of the target patient, compares the plurality of physiological feature data with known organ decline mode feature data, calculates the matching degree of the plurality of physiological feature data, evaluates the organ decline level of the patient, and generates organ decline monitoring information;

[0010] The progression stage identification module is based on the organ decline monitoring information, identifies the development stage of the patient's condition according to the blood glucose value, urine volume level and organ function level data of the patient, and generates a diabetes stage classification result.

[0011] As a further scheme of the present application, the blood glucose fluctuation pattern acquisition step is specifically:

[0012] Based on real-time blood glucose monitoring data, the formula:

[0013]

[0014] is used to calculate the blood glucose fluctuation degree score;

[0015] Wherein, x i is the blood glucose value measured for the i-th time, i represents the measurement index, N is the number of blood glucose measurements in a day, k is an adjustment coefficient for adjusting the sensitivity of the fluctuation score, and σ represents the daily blood glucose fluctuation degree score;

[0016] Based on the blood glucose fluctuation degree score, the blood glucose fluctuation degree of the target patient every day is analyzed, and the blood glucose fluctuation pattern is identified in combination with the fluctuation period of the patient's blood glucose data.

[0017] As a further scheme of the present application, the blood glucose feature data set acquisition step is specifically:

[0018] Based on the blood glucose fluctuation pattern, the dietary record information of the target patient is collected, the blood glucose measurement value corresponding to each of a plurality of dietary events and the corresponding nutrient intake are extracted, and the formula:

[0019]

[0020] is used to calculate the coefficient of the influence of nutrient intake on blood glucose,

[0021] Wherein, ΔX j is the difference between the blood glucose after eating and the previous measurement, C j is the corresponding nutrient intake, and α represents the influence degree of the nutrient intake level on the blood glucose change.j represents the blood glucose value measured after a meal, j is an index for marking a meal event;

[0022] based on the coefficient of the influence of the nutritional intake on blood glucose, analyzing blood glucose responses under multiple nutritional intake levels, evaluating expected blood glucose fluctuations under multiple meal situations, and generating a sensitivity analysis result;

[0023] According to the sensitivity analysis result, predicting the blood glucose level development trend of the patient according to the eating habits and nutritional intake level of the patient, and generating a blood glucose feature dataset.

[0024] As a further scheme of the present application, the obtaining step of the polyuria symptom level is specifically:

[0025] Based on the blood glucose feature dataset, collecting the drinking water and urination data of the patient, recording the urination frequency and urination volume of the patient, and generating patient urine volume data;

[0026] Based on the patient urine volume data, the severity score of the polyuria symptom is calculated by the formula:

[0027]

[0028]

[0029] wherein, U o is the urine volume of each urination, Ul is the ratio of actual urine volume to ideal urine volume, which is used to quantify the severity of polyuria symptoms, Ub is the basal urine volume, k a is the adjustment coefficient of age, k w is the adjustment coefficient of body weight, k s is the adjustment coefficient of gender, a is the actual age of the patient, w is the body weight of the patient, s is the gender of the patient, and o is the index of the urination event;

[0030] According to the severity score of the polyuria symptom, the polyuria symptom level of the target patient is classified, and the polyuria symptom level is generated.

[0031] As a further scheme of the present application, the obtaining step of the polyuria symptom analysis result is specifically:

[0032] Based on the polyuria symptom level, the daily urine volume data and blood glucose data of the target patient are extracted, and the correlation between the daily urine volume and the blood glucose data of the patient is calculated by the formula:

[0033]

[0034]

[0035] wherein, Y l represents the daily urine volume measured at a time, ​​S represents the average value of daily urine volume data, r represents the correlation coefficient, and l represents the index of the data point l S represents the average value of blood glucose data, r represents the correlation coefficient, and l represents the index of the data point S represents the average value of blood glucose data, r represents the correlation coefficient, and l represents the index of the data point

[0036] Based on the correlation coefficient, the correlation between the polydipsia and polyuria symptom level and the diabetes progression stage is evaluated, and the polyuria symptom analysis result is output in combination with the severity of the polydipsia and polyuria symptoms of the patient.

[0037] As a further scheme of the present application, the matching degree of the plurality of physiological characteristic data is specifically obtained by:

[0038] Based on the polyuria symptom analysis result, a plurality of physiological parameter data of a target patient is collected, including blood pressure, electrocardiogram, and renal function indicators, and the change trend and fluctuation mode of the plurality of parameters are calculated, including the average change rate and the fluctuation index, to generate data change characteristic values;

[0039] Based on the data change characteristic values, the similarity between the measured physiological data and the known organ decline mode characteristic data is calculated by the formula:

[0040]

[0041] The similarity between the measured physiological data and the known organ decline mode characteristic data is calculated by the formula:

[0042] S represents the average value of daily urine volume data, r represents the correlation coefficient, and l represents the index of the data point pq S represents the average value of daily urine volume data, r represents the correlation coefficient, and l represents the index of the data point p S represents the average value of daily urine volume data, r represents the correlation coefficient, and l represents the index of the data point p S represents the average value of daily urine volume data, r represents the correlation coefficient, and l represents the index of the data point q S represents the average value of daily urine volume data, r represents the correlation coefficient, and l represents the index of the data point q S represents the average value of daily urine volume data, r represents the correlation coefficient, and l represents the index of the data point

[0043] Based on the physiological characteristic similarity value, the matching degrees of the plurality of physiological characteristic data are generated by summarizing the similarity calculation results of the plurality of physiological parameter characteristic values.

[0044] As a further scheme of the present application, the organ decline monitoring information is specifically obtained by:

[0045] Based on the matching degrees of the plurality of physiological characteristic data, influence weights are assigned to the plurality of physiological parameters to reflect the actual influence of the plurality of physiological parameters on the health status of the patient, and characteristic weight matching degree data is generated;

[0046] Based on the characteristic weight matching degree data, the organ decline monitoring information is generated by the formula:

[0047]

[0048] calculate the degree of organ decline of the patient, obtain organ decline degree prediction data;

[0049] wherein R represents the predicted organ decline degree, β0 is the intercept of the regression model, β z is the regression coefficient of the zth physiological characteristic, Q z is the matching degree value of the zth physiological characteristic, z represents the index of a specific physiological characteristic, and Z represents the total number of physiological characteristics;

[0050] based on the organ decline degree prediction data, real-time evaluation of the organ decline level of the target diabetes evaluation patient is performed to generate organ decline monitoring information.

[0051] As a further scheme of the present application, the obtaining of the diabetes stage classification result specifically comprises:

[0052] based on the organ decline monitoring information, blood glucose value, urine volume level and organ function level data of the patient are extracted to generate diabetes progression feature data;

[0053] based on the diabetes progression feature data, the diabetes progression stage of the target patient is determined through the formula:

[0054] P′=a′1·D1+a′2·D2+a′3·D3;

[0055] calculate the condition score of the target patient;

[0056] P′ represents the comprehensive score of the condition, D1 represents the standardized score of the blood glucose value, D2 represents the standardized score of the urine volume, and D3 represents the standardized score of the organ function, a′1 is the weight coefficient of the blood glucose value, a′2 is the weight coefficient of the urine volume, and a′3 is the weight coefficient of the organ function;

[0057] based on the condition score of the target patient, in combination with the opinions of physicians, the condition development stage of the target diabetes patient is real-time evaluated to generate a diabetes stage classification result.

[0058] Compared with the prior art, the present application has the advantages and positive effects that:

[0059] In the application, through real-time blood glucose monitoring and correlation analysis of the patient's diet state, the change of blood glucose level is accurately captured, the development trend of the disease is predicted, the real-time of disease monitoring is enhanced, the medical intervention is more timely and effective, through continuous monitoring of the patient's water intake and urine output, the symptoms of polydipsia and polyuria of diabetic patients are found in time, and the correlation between the target symptoms and the progression of diabetes is determined through data analysis, the identification ability of early diabetic complications is enhanced, combined with the assessment of organ function level, the signs of organ damage are identified, the treatment strategy is adjusted, the development of complications is prevented, the real-time and accuracy of disease monitoring are improved, and the complexity and cost of long-term treatment are reduced. BRIEF DESCRIPTION OF DRAWINGS

[0060] Figure 1 is a system flowchart of the application;

[0061] Figure 2 is an analysis of blood glucose fluctuation mode flowchart of the application;

[0062] Figure 3 is a blood glucose feature data set flowchart of the application;

[0063] Figure 4 is a calculation of polyuria symptom level flowchart of the application;

[0064] Figure 5 is a generation of polyuria symptom analysis result flowchart of the application;

[0065] Figure 6 is a calculation of the matching degree of multiple physiological characteristic data flowchart of the application;

[0066] Figure 7 is an organ decline monitoring information flowchart of the application;

[0067] Figure 8 is a diabetes stage classification result flowchart of the application. DETAILED DESCRIPTION

[0068] In order to make the purpose, technical scheme and advantages of the application more clear and obvious, the application will be further described in detail below in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the application and do not limit the application.

[0069] In the description of the present application, it should be understood that the terms "length", "width", "upper", "lower", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship shown in the drawings, which are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation of the present application. In addition, in the description of the present application, the meaning of "a plurality of" is two or more, unless otherwise explicitly and specifically limited.

[0070] Referring to Figure 1 A diabetes symptom automatic detection system comprises:

[0071] The blood glucose fluctuation analysis module analyzes the blood glucose fluctuation pattern of the patient based on real-time blood glucose monitoring data, analyzes the influence of nutrient intake on blood glucose fluctuation in combination with the diet state of the patient, predicts the development trend of the blood glucose level of the patient, and generates a blood glucose feature data set;

[0072] The polydipsia and polyuria symptom detection module evaluates the polyuria symptom level of the patient by collecting the water intake and urine volume data of the patient based on the blood glucose feature data set, and calculates the correlation between the polyuria symptom feature and the progression stage of diabetes according to the correlation between the urine volume level and the blood glucose feature, and outputs the polyuria symptom analysis result;

[0073] The decline level evaluation module collects a plurality of physiological feature data of the target patient using the polyuria symptom analysis result, compares the plurality of physiological feature data with known organ decline pattern feature data, calculates the matching degree of the plurality of physiological feature data, evaluates the organ decline level of the patient, and generates organ decline monitoring information;

[0074] The progression stage identification module identifies the development stage of the patient's condition based on the blood glucose value, urine volume level and organ function level data of the patient according to the organ decline monitoring information, and generates a diabetes stage classification result.

[0075] The blood glucose feature data set is specifically the blood glucose peak and valley value, blood glucose fluctuation period, and nutrient intake influence analysis result, the polyuria symptom analysis result is specifically the polyuria symptom level, diabetes progression stage correlation index, and patient water intake and urine volume data, the organ decline monitoring information is specifically the organ decline level calculation result, data feature matching degree, and patient physiological feature data set, and the diabetes stage classification result is specifically the condition progression classification, patient condition risk score, and patient condition development trend information.

[0076] Referring to Figure 2 The blood glucose fluctuation pattern acquisition step is specifically:

[0077] Based on real-time blood glucose monitoring data, the formula:

[0078]

[0079] Calculate the blood glucose fluctuation degree score;

[0080] wherein x i is the blood glucose value of the ith measurement, i represents the measurement index, N is the number of blood glucose measurements in a day, k is an adjustment coefficient for adjusting the sensitivity of the fluctuation score, and σ represents the daily blood glucose fluctuation degree score.

[0081] Formula:

[0082]

[0083] Parameter meaning and acquisition method:

[0084] x i : the blood glucose value of the ith measurement, which is obtained in real time by a blood glucose monitoring device;

[0085] N: the number of blood glucose measurements in a day, which is obtained by setting the daily record of the blood glucose monitoring device;

[0086] k: adjustment coefficient, used to adjust the sensitivity of the fluctuation score;

[0087] σ: represents the daily blood glucose fluctuation degree score;

[0088] Calculation example:

[0089] Suppose there are 5 blood glucose measurements in total, i.e. N = 5, and the measurement values x i are 100, 105, 98, 103, and 110 respectively, and the adjustment coefficient k is 1.5. Calculate the blood glucose fluctuation degree:

[0090]

[0091]

[0092]

[0093] σ = 1.5 x 4.66;

[0094] σ = 6.99;

[0095] The calculation result σ = 6.99 indicates that the blood glucose fluctuation degree score of the patient for the day is 6.99, and the numerical value provides an enhanced evaluation of the sensitivity of blood glucose fluctuation, which is used to understand the volatility of the patient's blood glucose control.

[0096] Based on the blood glucose fluctuation degree score, analyze the daily blood glucose fluctuation degree of the target patient, and identify the blood glucose fluctuation pattern in combination with the fluctuation period of the patient's blood glucose data.

[0097] By collecting blood glucose monitoring data of the target patient within a certain period, using time series analysis method to analyze the data, determining the fluctuation frequency and amplitude of blood glucose, identifying the main mode of blood glucose fluctuation such as periodic peak and trough, through mathematical modeling technology, using Fourier transform analysis method, frequency domain conversion is carried out on time series data, the periodic characteristics of blood glucose data are extracted, which helps medical providers to deeply understand the fluctuation rule of the disease, provides basis for adjusting treatment plan, and the generated detailed blood glucose fluctuation report is used for optimizing disease management strategy and improving patient treatment response.

[0098] Please refer to Figure 3 The acquisition step of the blood glucose feature data set is specifically:

[0099] Based on the blood glucose fluctuation mode, the diet record information of the target patient is collected, the blood glucose measurement value corresponding to multiple diet events and the corresponding nutrient intake are extracted, and the formula:

[0100]

[0101] The coefficient of the influence of nutrient intake on blood glucose is calculated,

[0102] Where, ΔX j is the difference between the blood glucose after diet and the previous measurement, C j is the corresponding nutrient intake, and a represents the influence degree of the nutrient intake level on the change of blood glucose, X j represents the measured blood glucose value after diet, and j is the index for marking the diet event.

[0103] The formula is:

[0104]

[0105] Parameter meaning and acquisition method:

[0106] C j : The carbohydrate intake of the jth diet event is obtained through food intake recorded in the diet log and the nutrient database;

[0107] X j : The blood glucose value measured after the jth diet event is obtained through the blood glucose monitoring device;

[0108] ΔX j : The change amount of blood glucose between the jth diet and the previous measurement;

[0109] a: The coefficient of the influence of nutrient intake on blood glucose, used to evaluate the average influence of diet on blood glucose change;

[0110] Calculation example:

[0111] Set C jare 50, 60, 40, ΔX j are 20, 15, 10, calculate the impact coefficient:

[0112] Calculate ∑(C j × ΔX j ):

[0113] ∑(C j × ΔX j ) = 50 × 20 + 60 × 15 + 40 × 10 = 1000 + 900 + 400 = 2300; calculate

[0114]

[0115] Calculate α:

[0116]

[0117] α = 0.2987 indicates that each unit of carbohydrates causes an average change in blood glucose of 0.2987, the value reflects the average strength of the food on blood glucose, the coefficient is used to assess the effect of dietary adjustments on blood glucose control.

[0118] Based on the coefficient of the impact of nutritional intake on blood glucose, analyze the blood glucose response under various levels of nutritional intake, evaluate the expected blood glucose fluctuations under various dietary conditions, and generate sensitivity analysis results;

[0119] By analyzing the patient's past diet records and corresponding blood glucose response data, using regression analysis techniques to quantitatively assess the relationship between food intake and blood glucose response, in the analysis, determine the blood glucose impact coefficient of each type of food, calculate the correlation coefficient between nutrients and blood glucose changes, the sensitivity analysis results show the impact of food composition on blood glucose control, through dynamic simulation of blood glucose response under different dietary scenarios, construct a food blood glucose response model, provide personalized dietary recommendations for patients, guide patients to adjust their daily diet, and optimize blood glucose control strategies.

[0120] According to the sensitivity analysis results, according to the patient's eating habits and nutritional intake level, predict the patient's blood glucose level development trend, generate blood glucose feature data set;

[0121] Apply linear regression model in machine learning technology to predict the future trend of the patient's blood glucose level, collect and organize the patient's historical diet and blood glucose data, use data to train the prediction model, ensure that the model can accurately reflect the patient's blood glucose change rule, through algorithm adjustment and optimization, the generated blood glucose feature data set provides a scientific basis for medical decision-making, improves the accuracy of diabetes management and the quality of life of patients.

[0122] Please refer to Figure 4, the acquisition step of the polyuria symptom level is specifically:

[0123] Based on the blood glucose feature dataset, the patient's water intake and urine output data are collected, the patient's urination frequency and urination volume are recorded, and the patient's urine volume data are generated;

[0124] The automatic acquisition system is set to record the patient's water intake and subsequent urine output every hour, detailed water intake and urine output data are recorded using sensors and mobile applications, the collected data are quality checked to exclude any outliers such as abnormally high urine output or unrecorded water intake events, the daily total urine output and urine output fluctuations are calculated by statistical software, the intra-day and inter-day urine output changes are evaluated, and the patient's detailed urine output data are generated. After the data are processed by the algorithm, they are converted into trend charts and cycle charts that are easy to interpret, which assist medical professionals in monitoring and evaluating the patient's water metabolism health and changes in diabetes condition.

[0125] Based on the patient's urine volume data, the formula:

[0126]

[0127] The severity score of polyuria symptoms is calculated;

[0128] wherein U o is the urine output of each urination, Ul is the ratio of actual urine output to ideal urine output, which is used to quantify the severity of polyuria symptoms, Ub is the basal urine output, k a is the age adjustment coefficient, k w is the weight adjustment coefficient, and k s is the gender adjustment coefficient, a is the actual age of the patient, w is the weight of the patient, s is the gender of the patient, and o is the index of the urination event.

[0129] The formula:

[0130]

[0131] Parameter meaning and acquisition method

[0132] U o : the urine output of each urination of the patient, which is obtained by the hospital's urine output measurement equipment;

[0133] Ub: basal urine output, representing the average urine output standard of healthy adults, determined based on medical research;

[0134] k a : age adjustment coefficient, determined according to clinical research, used to adjust the urine output according to the patient's age;

[0135] k w : weight adjustment coefficient, determined according to clinical research, used to adjust the urine output according to the patient's weight;

[0136] k s : gender adjustment factor, based on research of gender differences;

[0137] a: actual age of the patient, obtained from the patient's medical records;

[0138] w: actual weight of the patient, obtained from the patient's medical records;

[0139] s: gender of the patient, obtained from the patient's medical records;

[0140] Calculation Example

[0141] Set the actual urine volume U of the target patient per time o Data is [500, 550, 520, 540] mL, Ub = 1500 mL / day, k a = 0.5 mL / day / year, k w = 1 mL / day / kg, k s = 100 mL / day, a = 55 years, w = 70 kg, s = 1, calculate the severity of symptoms:

[0142]

[0143] The result Ul = 1.243 indicates that the actual urine volume is 124.3% of the ideal urine volume, indicating that the patient has significant polyuria symptoms, and the numerical value is used to help doctors assess the condition and adjust treatment strategies.

[0144] According to the severity score of polyuria symptoms, classify the polyuria symptom level of the target patient, and generate the polyuria symptom level;

[0145] Compare the patient's urine volume data with the clinical symptom records, use a standardized scoring table to quantify the severity of each symptom, use statistical analysis tools such as factor analysis to determine the contribution of each symptom to the total score, objectively classify the patient's polyuria symptoms, and generate a polyuria symptom level report for each patient based on the score results and relevant medical guidelines. This report lists the specific manifestations of the patient's polyuria symptoms and recommended medical responses, providing a basis for developing targeted treatment plans.

[0146] Please refer to Figure 5 , the steps to obtain the polyuria symptom analysis results are as follows:

[0147] Based on the polyuria symptom level, extract the daily urine volume data and blood glucose data of the target patient, and use the formula:

[0148]

[0149] Correlation between daily urine volume and patient blood glucose data is calculated, and a correlation coefficient is obtained;

[0150] Y l represents the daily urine volume of a single measurement, represents the average value of daily urine volume data, S l represents the blood glucose value of a single measurement, represents the average value of blood glucose data, and r represents the correlation coefficient, which measures the strength of linear correlation between daily urine volume Y and blood glucose S, and I represents the index of data points.

[0151] Formula:

[0152]

[0153] Parameter meaning and acquisition method:

[0154] Y l : specific measurement value of daily urine volume;

[0155] average value of daily urine volume;

[0156] S l : specific measurement value of blood glucose;

[0157] average value of blood glucose data;

[0158] r: correlation coefficient, which measures the strength of linear correlation between daily urine volume Y and blood glucose S;

[0159] Calculation example

[0160] Set the daily urine volume data of the target patient as [500, 600, 550, 580, 610], The blood glucose data is [5.0, 5.5, 5.3, 5.2, 5.6], Calculate the correlation coefficient:

[0161]

[0162]

[0163]

[0164]

[0165] r≈0.901;

[0166] The calculation result r=0.901 indicates a positive correlation between urine volume and blood glucose, meaning that as blood glucose increases, urine volume also increases, reflecting the close correlation between polyuria symptoms and the progression stage of diabetes.

[0167] Based on the correlation coefficient, the correlation between the level of polydipsia and polyuria symptoms and the progression stage of diabetes is evaluated, and the severity of the patient's polydipsia and polyuria symptoms is combined to output the polyuria symptom analysis result;

[0168] The correlation between the level of polydipsia and polyuria symptoms and the progression stage of diabetes is evaluated using multivariate regression analysis, and the patient's urine volume data and blood glucose monitoring results are integrated to apply statistical model analysis to the target data, identify potential correlation patterns, and output detailed polyuria symptom analysis results that provide important insights into the patient's disease progression, support clinical decision-making, and help doctors adjust treatment plans for patients, including in diabetes management and complication prevention.

[0169] Please refer to Figure 6 , the matching degree of the plurality of physiological characteristic data is obtained by:

[0170] Based on the polyuria symptom analysis result, a plurality of physiological parameter data of the target patient is collected, including blood pressure, electrocardiogram, and renal function indicators, and the change trend and fluctuation pattern of the plurality of parameters are calculated, including the average change rate and the fluctuation index, and the data change characteristic value is generated;

[0171] The comprehensive collection of the target patient's physiological parameters is carried out, covering the data collected by sphygmomanometer, electrocardiograph and renal function test equipment, and each device is calibrated to ensure data accuracy. The patient's daily blood pressure, electrocardiogram and renal function indicators are systematically recorded, and the change trend and fluctuation of each physiological parameter are evaluated through data analysis software using sliding window average method and standard deviation calculation method to identify key health change signals and calculate data change characteristic values including average change rate and fluctuation index. The characteristic values serve as an important basis for evaluating the patient's health status and disease changes.

[0172] Based on the data change characteristic value, the similarity between the measured physiological data and the known organ decline pattern characteristic data is calculated by the formula:

[0173]

[0174] The similarity between the measured physiological data and the known organ decline pattern characteristic data is calculated by the formula:

[0175] Where S pq is the similarity index between the measured data and the standard pattern, V p is the average change rate of the measured data, F p is the fluctuation index of the measured data, and Vq V is the average change rate of the known standard data, F q p represents the index of the measured data characteristics, including blood pressure, electrocardiogram, and renal function level, q represents the index of the known organ decline pattern characteristic data;

[0176] Formula:

[0177]

[0178] Parameter meaning and acquisition method:

[0179] V p : the average change rate of the measured physiological data;

[0180] F p : the fluctuation index of the measured physiological data;

[0181] V q : the average change rate of the known organ decline pattern data;

[0182] F q : the fluctuation index of the known organ decline pattern data:

[0183] Calculation example:

[0184] Suppose the measured physiological data points are [100, 105, 95, 110, 90], and the known organ decline pattern data points are [95, 100, 105, 95, 100], V p = -2.5, V q = 1.25, F p = 7.35, F q = 3.74, the similarity is calculated as:

[0185]

[0186]

[0187] The calculation result shows that the similarity between the measured value and the known organ decline pattern characteristic data is 0.795, which reflects the high similarity between the measured data and the known decline pattern. The calculation process is used to evaluate the matching degree of multiple physiological characteristics and pathological models.

[0188] Based on the physiological characteristic similarity value, the matching degrees of multiple physiological characteristic data are generated by summarizing the similarity calculation results of multiple physiological parameter characteristic values;

[0189] The physiological parameter characteristic values are calculated for similarity using multivariate data analysis techniques such as principal component analysis and cluster analysis, the physiological parameters are standardized to eliminate the dimension effect, PCA is applied to extract the main change trend and mode, the similarity between the parameters is further determined through cluster analysis, the correlation between different physiological parameters is revealed in the calculation process, which provides a scientific basis for comprehensive judgment of the health status of the patient, the similarity calculation results are summarized to generate comprehensive multi-physiological characteristic data matching degree, which provides accurate health assessment and disease monitoring tools for medical professionals.

[0190] Please refer to Figure 7 The organ decline monitoring information acquisition step is specifically:

[0191] Based on the matching degree of multiple physiological characteristic data, the influence weight of multiple physiological parameters is allocated to reflect the actual influence of multiple physiological parameters on the health status of the patient, and characteristic weight matching degree data is generated;

[0192] A weight allocation model such as weighted linear regression analysis is used to allocate corresponding influence weights for different physiological parameters such as blood pressure, electrocardiogram and renal function indicators, a baseline influence model of each parameter on the health status of the patient is established through historical data, the weight of each parameter is adjusted according to the current matching degree data to reflect its actual influence in the current health status evaluation, the characteristic weight of each physiological parameter is calculated, and the comprehensive characteristic weight matching degree data is generated by summarizing the weights, which helps medical professionals understand the importance of each physiological indicator on the disease, and optimizes the treatment and monitoring strategy.

[0193] Based on the characteristic weight matching degree data, the organ decline degree of the patient is calculated through the formula:

[0194]

[0195] The organ decline degree prediction data is obtained;

[0196] Wherein, R represents the predicted organ decline degree, β0 is the intercept of the regression model, β z is the regression coefficient of the zth physiological characteristic, Q z is the matching degree value of the zth physiological characteristic, z represents the index of a specific physiological characteristic, and Z represents the total number of physiological characteristics;

[0197] The formula is:

[0198]

[0199] Parameter meaning and acquisition method:

[0200] R: predicted organ decline degree;

[0201] β0: intercept of the regression model, obtained by analyzing historical health data, representing the basic decline level without any physiological parameter influence;

[0202] β z : regression coefficient of the zth physiological parameter, reflecting the influence weight of the physiological parameter on the degree of decline;

[0203] Q z : matching degree value of the zth physiological parameter, obtained from physiological characteristic matching degree data, used as input for linear regression analysis;

[0204] z: physiological parameter index;

[0205] Z: total number of physiological parameters;

[0206] Calculation example:

[0207] Set Q1 = 0.8, Q2 = 0.6, Q3 = 0.9, β0 = 0.5, β1 = 0.3, β2 = 0.2, β3 = 0.4, Z = 3, calculate the degree of organ decline:

[0208] R = 0.5 + (0.3 * 0.8 + 0.2 * 0.6 + 0.4 * 0.9);

[0209] R = 0.5 + (0.24 + 0.12 + 0.36);

[0210] R = 0.5 + 0.72;

[0211] R = 1.22;

[0212] The calculation result shows that the predicted degree of organ decline is 1.22, which reflects the comprehensive decline state based on the given physiological parameter matching degree. The formula is used to evaluate the organ decline level of the patient, helping doctors make diagnosis and treatment decisions.

[0213] Based on the organ decline degree prediction data, real-time evaluation of the organ decline level of the target diabetes evaluation patient is performed to generate organ decline monitoring information;

[0214] Using dynamic models such as dynamic system models, combining the latest physiological parameter data of the patient and historical decline trends, real-time data analysis and prediction are performed, and through the model, the decline rate and degree of organ function are calculated and updated in real time, generating detailed organ decline monitoring information. The information provides immediate clinical data for doctors to support more accurate treatment decisions, while providing important information about the patient's disease progression, which helps to adjust personal health management plans and prevent potential complications.

[0215] Please refer to Figure 8 , the steps for obtaining the diabetes stage classification result are as follows:

[0216] Based on the organ decline monitoring information, the blood glucose value, urine volume level and organ function level data of the patient are extracted, and the diabetes progression feature data is generated;

[0217] The key physiological indicators are extracted by using data extraction technology, including the blood glucose value, urine volume level and organ function level data of the patient, the collected blood glucose and urine volume monitoring results are integrated by using an automatic data processing system, and the latest organ function detection report of the patient is synchronized, through data integration analysis, the change trend of multiple indicators in time sequence is evaluated, the time series analysis technology such as moving average or exponential smoothing technology is used to determine the fluctuation mode of blood glucose and urine volume and its correlation with organ function decline, and the process generates diabetes progression feature data, which provides accurate basic information for subsequent disease analysis and stage classification.

[0218] Based on the diabetes progression feature data, through the formula:

[0219] P′=a′1·D1+a′2·D2+a′3·D3;

[0220] The disease score of the target patient is calculated;

[0221] P′ represents the comprehensive score of the disease, D1 represents the standardized score of the blood glucose value, D2 represents the standardized score of the urine volume, and D3 represents the standardized score of the organ function, a′1 is the weight coefficient of the blood glucose value, a′2 is the weight coefficient of the urine volume, and a′3 is the weight coefficient of the organ function;

[0222] The formula is:

[0223] P′=a′1·D1+a′2·D2+a′3·D3;

[0224] Parameter meaning and acquisition method:

[0225] P′ represents the comprehensive score of the disease;

[0226] D1 represents the standardized score of the blood glucose value, which is obtained by standardizing the blood glucose data obtained by actual medical detection;

[0227] D2 represents the standardized score of the urine volume, which is obtained by standardizing the urine volume data obtained by actual medical detection;

[0228] D3 represents the standardized score of the organ function, which is obtained by standardizing the organ function test data obtained by actual medical detection;

[0229] a′1 is the weight coefficient of the blood glucose value, which is obtained by analyzing historical health data, reflecting the influence of blood glucose value in disease assessment;

[0230] a'2 is the weight coefficient of urine volume, obtained by analyzing historical health data, reflecting the influence of urine volume in the disease assessment;

[0231] a'3 is the weight coefficient of organ function, obtained by analyzing historical health data, reflecting the influence of organ function in the disease assessment;

[0232] Calculation example:

[0233] Set D1=0.85, D2=0.75, D3=0.65, a'1=0.4, a'2=0.3, a'3=0.3, calculate the comprehensive score of the disease:

[0234] P'=0.4*0.85+0.3*0.75+0.3*0.65;

[0235] P'=0.34+0.225+0.195;

[0236] P'=0.76;

[0237] The calculation result shows that the comprehensive score of the target patient's disease is 0.76, which reflects the disease development stage assessment score obtained after the comprehensive blood glucose value, urine volume and organ function, and the value is used to help the doctor determine the disease stage of the diabetic patient, and provide basis for formulating treatment plan.

[0238] Based on the disease score of the target patient, combined with the doctor's opinion, the disease development stage of the target diabetic patient is assessed in real time, and the diabetic stage classification result is generated;

[0239] By integrating the scores and opinions provided by the doctors, using a decision support system, combining logistic regression analysis, the blood glucose control, urine volume change and organ function indicators of the patient are comprehensively evaluated, the disease development stage of the diabetic patient is assessed in real time, the model calculation considers the correlation between the physiological data and the progression of diabetes and the influence on the disease stage, and the output diabetic stage classification result provides clear information about the current stage of the disease for the doctor and the patient, supporting personalized treatment decisions and disease management strategies.

[0240] The above is only a preferred embodiment of the present application, and does not limit the form of the present application, any skilled person in the art can use the disclosed technical content to make changes or modifications as equivalent embodiments applied to other fields, but any simple modification, equivalent change and modification made according to the technical essence of the present application to the above embodiments, still belongs to the protection scope of the technical solution of the present application.

Claims

1. An automatic detection system of diabetic symptoms, characterized by, The system comprises: The blood glucose fluctuation analysis module analyzes the blood glucose fluctuation pattern of the patient based on real-time blood glucose monitoring data, analyzes the influence of nutritional intake on blood glucose fluctuation in combination with the diet state of the patient, predicts the development trend of the blood glucose level of the patient, and generates a blood glucose feature data set; The polydipsia and polyuria symptom detection module evaluates the polyuria symptom level of the patient by collecting the water intake and urine volume data of the patient based on the blood glucose feature data set, and calculates the correlation between the polyuria symptom feature and the progression stage of diabetes according to the correlation between the urine volume level and the blood glucose feature, and outputs the polyuria symptom analysis result; The decline level evaluation module collects a plurality of physiological feature data of the target patient by using the polyuria symptom analysis result, compares with the known organ decline pattern feature data, calculates the matching degree of the plurality of physiological feature data, evaluates the organ decline level of the patient, and generates organ decline monitoring information; The progression stage identification module identifies the development stage of the patient's condition according to the blood glucose value, urine volume level and organ function level data of the patient based on the organ decline monitoring information, and generates a diabetes stage classification result; The blood glucose feature data set is obtained by: Based on the blood glucose fluctuation pattern, the diet record information of the target patient is collected, the blood glucose measurement value corresponding to each diet event and the corresponding nutritional intake are extracted, and the coefficient of the influence of nutritional intake on blood glucose is calculated by the formula: ； Based on the coefficient of the influence of nutritional intake on blood glucose, the blood glucose response under a plurality of nutritional intake levels is analyzed, the expected blood glucose fluctuation under a plurality of diet conditions is evaluated, and a sensitivity analysis result is generated; wherein, is the difference between the post-prandial blood glucose and the previous measurement, is the corresponding nutritional intake, represents the degree of influence of the nutritional intake level on the blood glucose change, represents the blood glucose value measured post-prandially, is an index used to mark a meal event; According to the sensitivity analysis result, the development trend of the blood glucose level of the patient is predicted according to the diet habit and nutritional intake level of the patient, and a blood glucose feature data set is generated; The polyuria symptom level is obtained by: Based on the blood glucose feature data set, the water intake and urination data of the patient are collected, the urination frequency and urination volume of the patient are recorded, and the urine volume data of the patient is generated; Based on the urine volume data of the patient, the severity score of the polyuria symptom is calculated by the formula: According to the severity score of the polyuria symptom, the polyuria symptom level of the target patient is classified, and a polyuria symptom level is generated; ； The polyuria symptom analysis result is obtained by: wherein, is the urine volume for each micturition, is the ratio of the actual urine volume to the ideal urine volume, quantifying the severity of the polyuria symptom, is the basal urine volume, is the adjustment coefficient for the age, is the adjustment coefficient for the weight, is the adjustment coefficient for the gender, is the actual age of the patient, is the weight of the patient, is the gender of the patient, is the index of the micturition event; Based on the polyuria symptom level, the daily urine volume data and blood glucose data of the target patient are extracted, the correlation between the daily urine volume and the blood glucose data of the patient is calculated by the formula: Based on the correlation coefficient, the correlation between the polydipsia and polyuria symptom level and the progression stage of diabetes is evaluated, and the polyuria symptom analysis result is output in combination with the severity of the polydipsia and polyuria symptom of the patient; The matching degree of the plurality of physiological feature data is obtained by: ； Based on the polyuria symptom analysis result, a plurality of physiological parameter data of the target patient are collected, including blood pressure, electrocardiogram and renal function indicators, the change trend and fluctuation pattern of the plurality of parameters are calculated, including the average change rate and the fluctuation index, and data change characteristic values are generated; wherein, represents the daily urine volume of a single measurement, represents the mean value of the daily urine volume data, represents the blood glucose value of a single measurement, represents the mean value of the blood glucose data, represents the correlation coefficient, represents the index of the data point; Based on the data change characteristic values, the similarity between the measured physiological data and the known organ decline pattern feature data is calculated by the formula: A physiological feature similarity value is obtained. ​ ​ ； ​ wherein, is an index of similarity between the measured data and the standard pattern, is an average rate of change of the measured data, is a fluctuation index of the measured data, is an average rate of change of the known standard data, is a fluctuation index of the known standard data, is an index representing a characteristic of the measured data, is an index representing a characteristic of the known organ decline pattern data; Based on the physiological characteristic similarity value, the matching degrees of the multiple physiological characteristic data are generated by aggregating the similarity calculation results of the multiple physiological parameter characteristic values; The organ decline monitoring information acquisition step is specifically: Based on the matching degrees of the multiple physiological characteristic data, the influence weights of the multiple physiological parameters are assigned to reflect the actual influence of the multiple physiological parameters on the health status of the patient, and the characteristic weight matching degree data is generated; Based on the characteristic weight matching degree data, the organ decline degree of the patient is calculated by the formula: ； to obtain the organ decline degree prediction data; wherein, represents a predicted degree of organ deterioration, is an intercept of the regression model, is a regression coefficient of the zth physiological characteristic, is a matching degree value of the zth physiological characteristic, represents an index of a specific physiological characteristic, represents a total number of physiological characteristics; Based on the organ decline degree prediction data, the organ decline level of the target diabetic patient is evaluated in real time to generate the organ decline monitoring information.

2. The automatic detection system of diabetes symptoms according to claim 1, characterized in that, The blood glucose fluctuation pattern acquisition step is specifically: Based on the real-time blood glucose monitoring data, the blood glucose fluctuation degree score is calculated by the formula: ； ; wherein, is the blood glucose value of the first measurement, is the blood glucose value of the second measurement, represents the measurement index, is the number of blood glucose measurements in a day, is an adjustment factor to adjust the sensitivity of the fluctuation score, denotes the daily blood glucose fluctuation score; Based on the blood glucose fluctuation degree score, the blood glucose fluctuation degree of the target patient per day is analyzed, and the blood glucose fluctuation pattern is identified in combination with the fluctuation period of the blood glucose data of the patient.

3. The automatic detection system of diabetes symptoms according to claim 1, characterized in that, The diabetes stage classification result acquisition step is specifically: Based on the organ decline monitoring information, the blood glucose value, urine volume level and organ function level data of the patient are extracted to generate the diabetes progression characteristic data; Based on the diabetes progression characteristic data, the disease score of the target patient is calculated by the formula: ； ; a composite score indicative of the condition, a standardized score indicative of the blood glucose value, a standardized score indicative of the urine volume, a standardized score indicative of the organ function, is a weight coefficient for the blood glucose value, is a weight coefficient for the urine volume, is a weight coefficient for the organ function; Based on the disease score of the target patient, the disease development stage of the target diabetic patient is evaluated in real time in combination with the physician's opinion to generate the diabetes stage classification result.

Images