A method and system for assessing risk of diabetes

Abstract

The present application relates to the technical field of diabetes risk assessment, and discloses a diabetes risk assessment method and system, comprising: collecting user data, obtaining a diabetes diagnosis standard, and performing crowd determination according to the user data and the diabetes diagnosis standard; establishing a life correction model, a glycation deviation model and a glycosylation accumulation model, processing the user data to obtain analysis data; when the crowd determination is a non-diabetes crowd, establishing an onset evaluation model, the onset evaluation model calculating a basic risk score based on the life correction model, correcting the basic risk score through onset condition auditing to obtain a glycosylation corrected risk score, and mapping the glycosylation corrected risk score to an onset risk grade; when the crowd determination is a diabetes crowd, establishing a control evaluation model, the control evaluation model calculating a basic control risk grade based on the life correction model and the glycosylation accumulation model, and correcting the basic control risk grade through control condition auditing to obtain a control risk grade.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of diabetes risk assessment, in particular to a diabetes risk assessment method and system. BACKGROUND

[0002] Diabetes is a metabolic disease characterized by chronic hyperglycemia, and its onset process is occult and long-term. In the prior art, the risk assessment for diabetes mostly adopts a single population assessment method, that is, the onset risk is predicted for non-diabetic population respectively, or the blood glucose control condition is analyzed for diabetic patients. However, in the actual health management process, there are often two types of population in the same system: one type is individuals who have not been diagnosed but have potential risk of onset, and the other type is diabetic patients who have been diagnosed and treated. The existing method cannot assess the health risks of the two types of population in a unified framework, resulting in scattered assessment results, poor data continuity, and difficulty in maintaining consistency in long-term health monitoring.

[0003] At the same time, the traditional risk assessment method generally relies on single-point blood glucose or single glycated hemoglobin detection results, and does not consider key factors such as blood glucose fluctuation, individual glycation difference and long-term high blood glucose exposure, thereby there is a significant deviation among different individuals, and it is difficult to accurately reflect the true risk level. In addition, for the control assessment of diabetic patients, the current method mostly uses static indicators (such as HbA1c compliance rate) for judgment, lacks a dynamic analysis mechanism combining blood glucose time series data and glycation physiological characteristics, and is difficult to realize individualized quantitative assessment of risk.

[0004] In summary, the prior art urgently needs a unified method that can identify the types of population and respectively assess the onset risk and control risk, in order to realize continuous and quantifiable risk assessment for non-diabetic population and diabetic population. SUMMARY

[0005] In view of the above problems, the present application is proposed.

[0006] To solve the above technical problems, the present application provides the following technical scheme: a diabetes risk assessment method, comprising: collecting user data, obtaining a diabetes diagnosis standard, and performing population determination according to the user data and the diabetes diagnosis standard;

[0007] Establishing a life correction model, a glycation deviation model and a glycosylation accumulation model, processing the user data to obtain analysis data;

[0008] When the population determination is a non-diabetic population, an onset assessment model is established, the onset assessment model calculates a basic risk score based on the life correction model, modifies the basic risk score through onset condition audit to obtain a glycosylation modified risk score, and maps the glycosylation modified risk score to an onset risk grade;

[0009] When the crowd is determined as a diabetes crowd, a control evaluation model is established, the control evaluation model calculates a basic control risk level based on a life correction model and a glycation accumulation model, and the basic control risk level is corrected by a control condition audit to obtain a control risk level.

[0010] As a preferred scheme of the diabetes risk evaluation method, the user data is collected, a diabetes diagnosis standard is obtained, and crowd determination is performed according to the user data and the diabetes diagnosis standard, wherein the user data includes glycated hemoglobin data, fasting blood glucose data and blood glucose time series data.

[0011] The diabetes diagnosis standard is a numerical threshold for diagnosing diabetes, and the user data is compared with the diabetes diagnosis standard to perform crowd determination.

[0012] When the fasting blood glucose data is ≥7.0 mmol / L or the glycated hemoglobin data is ≥6.5%, the crowd is determined as a diabetes crowd.

[0013] When the fasting blood glucose data is <7.0 mmol / L and the glycated hemoglobin data is <6.5%, the crowd is determined as a non-diabetes crowd.

[0014] As a preferred scheme of the diabetes risk evaluation method, the life correction model is established, including calculating a glycation rate coefficient according to average blood glucose, calculating an equivalent red blood cell life according to the glycated hemoglobin data and the glycation rate coefficient, correcting the glycated hemoglobin data based on the equivalent red blood cell life, and outputting the corrected glycated hemoglobin data.

[0015] The glycation deviation model is established, including calculating expected glycated hemoglobin data according to average blood glucose, subtracting the glycated hemoglobin data from the glycated hemoglobin data, and outputting a glycation deviation value.

[0016] The glycation accumulation model is established, including setting a high blood glucose threshold of 7.8 mmol / L according to the diabetes diagnosis standard, and calculating a threshold exposure integral according to the blood glucose time series data.

[0017] ;

[0018] Wherein, E represents the threshold exposure integral; N represents the total number of blood glucose time series data; i represents the i-th data in the blood glucose time series data; max represents the maximum value. represents the blood glucose value in the i-th blood glucose time series data; represents the blood glucose value in the i+1-th blood glucose time series data; represents and the time difference between and

[0019] The average super-threshold exposure intensity is calculated by normalizing the super-threshold exposure integral to the full time length, and the glycation cumulative amount is obtained by linearly synthesizing the average super-threshold exposure intensity and the glycosylated hemoglobin data.

[0020] The analysis data includes the corrected glycosylated hemoglobin data, the glycation deviation value and the glycation cumulative amount.

[0021] As a preferred scheme of the diabetes risk assessment method, the disease incidence assessment model is based on a life correction model to calculate a basic risk score, the basic risk score is corrected through disease incidence condition auditing to obtain a glycation corrected risk score, and the glycation corrected risk score is mapped to a disease incidence risk level, which includes obtaining corrected glycosylated hemoglobin data according to the life correction model, and obtaining a glycation deviation value according to a glycation deviation model.

[0022] The corrected glycosylated hemoglobin data is taken as the basic input of the disease incidence assessment model, and the corrected glycosylated hemoglobin data is mapped to 0 to 1 by using a Sigmoid function to obtain a basic risk score.

[0023] The basic risk score is corrected through disease incidence condition auditing.

[0024] The basic risk score is corrected according to the glycation deviation value, if the absolute value of the glycation deviation value is greater than or equal to 0.6%, the glycation deviation value is called, the basic risk score is corrected by an exponential function, and the glycation corrected risk score is output, if the absolute value of the glycation deviation value is less than 0.6%, no correction is performed, and the glycation corrected risk score is equal to the basic risk score.

[0025] The glycation corrected risk score is corrected according to the glycation cumulative amount, when any glycation correction condition is met, the glycation cumulative amount is compressed by a Sigmoid function to obtain an intensity score, a gating coefficient is calculated according to the glycation corrected risk score, the glycation corrected risk score is calculated and output based on the gating coefficient and the intensity score, if the glycation correction condition is not met, no correction is performed, and the glycation corrected risk score is equal to the glycation corrected risk score.

[0026] The glycation correction condition includes two conditions.

[0027] The continuous time length of the blood glucose time series data greater than 7.8 mmol / L is 30 minutes;

[0028] The cumulative time length of the blood glucose time series data greater than 7.8 mmol / L in any 24 hours is 120 minutes;

[0029] The glycation corrected risk score is mapped to a disease incidence risk level.

[0030] As a preferred scheme of the diabetes risk assessment method, wherein: the control evaluation model is based on a life correction model and a glycation accumulation model to calculate a basic control risk level, and the basic control risk level is modified by a control condition audit to obtain a control risk level, which comprises: obtaining corrected glycosylated hemoglobin data according to the life correction model, and obtaining a glycation accumulation amount according to the glycation accumulation model;

[0031] Setting glycosylated hemoglobin data threshold and glycation accumulation amount threshold according to the diabetes diagnosis standard, mapping the corrected glycosylated hemoglobin data to a long-term control level according to the glycosylated hemoglobin data threshold, and mapping the glycation accumulation amount to a cumulative exposure level according to the glycation accumulation amount threshold, wherein the long-term control level and the cumulative exposure level are both divided into three levels of low, medium and high, and the higher level of the long-term control level and the cumulative exposure level is taken as the basic control risk level;

[0032] The basic control risk level is modified by a control condition audit:

[0033] When the absolute value of the glycation deviation value is greater than or equal to 0.6%, the basic control risk level is adjusted according to the positive and negative directions of the glycation deviation value, if the glycation deviation value is positive, the basic control risk level is adjusted by one level to obtain the control risk level, if the glycation deviation value is negative, the basic control risk level remains unchanged, and the control risk level is equal to the basic control risk level;

[0034] If the absolute value of the glycation deviation value is less than 0.6%, no modification is performed, and the control risk level is equal to the basic control risk level.

[0035] A diabetes risk assessment system using any of the methods described in the present application, wherein: a collection module collects user data, obtains a diabetes diagnosis standard, and performs population determination according to the user data and the diabetes diagnosis standard;

[0036] A calculation module establishes a life correction model, a glycation deviation model and a glycation accumulation model, processes the user data to obtain analysis data;

[0037] An onset evaluation module, when the population determination is a non-diabetes population, establishes an onset evaluation model, the onset evaluation model is based on the life correction model to calculate a basic risk score, the basic risk score is modified by an onset condition audit to obtain a glycation modified risk score, and the glycation modified risk score is mapped to an onset risk level;

[0038] A control evaluation module, when the population determination is a diabetes population, establishes a control evaluation model, the control evaluation model is based on the life correction model and the glycation accumulation model to calculate a basic control risk level, and the basic control risk level is modified by a control condition audit to obtain a control risk level.

[0039] The application provides a diabetes risk assessment method, and a multi-model fusion assessment system composed of a life correction model, a glycation deviation model and a glycosylation accumulation model is constructed, and unified modeling of first onset risk assessment of a non-diabetes population and control risk assessment of a diabetes population is realized. BRIEF DESCRIPTION OF DRAWINGS

[0040] In order to more clearly illustrate the technical solutions of the embodiments of the application, the drawings needed to be used in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative labor.

[0041] Figure 1 A diabetes risk assessment method provided for the embodiment 1 of the application is shown in the whole flowchart. DETAILED DESCRIPTION

[0042] In order to make the above-mentioned purposes, features and advantages of the application more obvious and easy to understand, the specific embodiments of the application will be described in detail below with reference to the drawings. Obviously, the described embodiments are part of the embodiments of the application, rather than all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor should belong to the protection scope of the application.

[0043] Embodiment 1, refer to Figure 1 For an embodiment of the application, a diabetes risk assessment method is provided, which comprises:

[0044] S1: collecting user data, obtaining a diabetes diagnosis standard, and performing population determination according to the user data and the diabetes diagnosis standard.

[0045] Further, user data is collected and diabetes diagnosis criteria is obtained. The user data includes glycated hemoglobin data, fasting blood glucose data and blood glucose time series data. The diabetes diagnosis criteria is a numerical threshold for diagnosing diabetes. The diabetes diagnosis criteria is an internationally recognized medical diagnosis standard, and the data is derived from the World Health Organization, which is a commonly used numerical threshold for diagnosing diabetes in the art and is directly applied in the present application.

[0046] Specifically, the glycated hemoglobin data is obtained by one test. The test can be completed in a medical institution or using a registered glycated hemoglobin test device. The test value and test date are recorded, and the unit is percentage.

[0047] The fasting blood glucose data is obtained under the condition that the fasting time before blood sampling is not less than 8 hours. The test can use intravenous blood enzyme method or fingertip blood glucose measurement using a portable blood glucose meter. The test value and test date are recorded, and the unit is mmol / L.

[0048] The blood glucose time series data is used for subsequent model calculation in the present application and is not used as a basis for population determination in this step. The data is collected in one of the following two ways, and either way is selected for implementation:

[0049] Continuous blood glucose meter test mode: install a continuous blood glucose monitoring device, set a fixed sampling interval, and continuously record blood glucose values from the installation time. Each record contains the date, time and blood glucose value, and the unit is mmol / L, forming blood glucose time series data.

[0050] Point measurement mode: use a portable blood glucose meter to measure fingertip blood glucose at a fixed time each day. Measure before breakfast, lunch, dinner and bedtime. Record the date, time and blood glucose value each time, and the unit is mmol / L. Sort the measurement results by time to form blood glucose time series data.

[0051] Further, by comparing the user data with the diabetes diagnosis criteria, the population determination is completed:

[0052] When the fasting blood glucose data is ≥7.0 mmol / L or the glycated hemoglobin data is ≥6.5%, the population is determined as a diabetes population;

[0053] When the fasting blood glucose data is <7.0 mmol / L and the glycated hemoglobin data is <6.5%, the population is determined as a non-diabetes population.

[0054] The above-mentioned various portable devices can independently complete the test in a home environment without relying on laboratory conditions of medical institutions.

[0055] S2: Establishing a life correction model, a glycation deviation model and a glycosylation accumulation model, processing user data and obtaining analysis data.

[0056] According to the blood glucose time series data, the blood glucose values in the blood glucose time series data are arithmetically averaged to obtain average blood glucose.

[0057] The life correction model is used to eliminate the influence of red blood cell life difference on glycated hemoglobin data. Specifically, the formation of glycated hemoglobin is determined by blood glucose level and red blood cell survival time. For each person, the red blood cell survival time basically does not change and can be regarded as a fixed value, so the formation rate of glycated hemoglobin is a certain value under the determined average blood glucose.

[0058] First, the glycation rate coefficient is calculated according to the average blood glucose:

[0059] ;

[0060] wherein, represents the glycation rate coefficient; , represents the glycation kinetics linear coefficient, a constant; represents the average blood glucose; the glycation rate coefficient is used to represent the glycation generation amount per unit time under the average blood glucose level.

[0061] Then, the glycated hemoglobin data is matched with the glycation rate coefficient to obtain the equivalent red blood cell life matched with the glycated hemoglobin data:

[0062] ;

[0063] wherein, L represents the equivalent red blood cell life; represents the glycated hemoglobin data.

[0064] Finally, the glycated hemoglobin data is standardized and converted by the reference red blood cell life to obtain the corrected glycated hemoglobin data:

[0065] ;

[0066] wherein, represents the corrected glycated hemoglobin data; represents the reference red blood cell life, i.e. 120 days. The corrected glycated hemoglobin data ensures that the subsequent calculation is carried out under the condition of uniform life, avoiding systematic deviation caused by individual red blood cell life difference.

[0067] It should be noted that the reference red blood cell lifespan refers to the average physiological lifespan of red blood cells from generation to clearance in healthy adult non-anemic population. In the present application, the reference red blood cell lifespan is used to standardize the glycated hemoglobin results of different individuals, i.e. to convert the measured glycated value to the glycated level when the assumed red blood cell lifespan is normal, so as to eliminate the systematic deviation caused by individual lifespan differences. The average lifespan of red blood cells in healthy adults is about 100-120 days, and the reference red blood cell lifespan is set to 120 days in the present embodiment.

[0068] The glycated deviation model is used to quantify whether the glycated hemoglobin data deviates from its expected level under a given average blood glucose. There is a stable one-to-one correspondence between the average blood glucose and the glycated hemoglobin data, and the expected glycated hemoglobin data can be obtained from the regression relationship. Specifically, the average blood glucose unit is converted from mmol / L to mg / dL, and the model first obtains the expected glycated hemoglobin data according to the average blood glucose:

[0069] ;

[0070] wherein, represents the expected glycated hemoglobin data; represents the average blood glucose after unit conversion; and the glycated deviation value is obtained by subtracting the expected glycated hemoglobin data from the measured glycated hemoglobin data.

[0071] The glycated deviation value is a quantitative result: a positive value indicates that the glycation is higher under the same average blood glucose, and a negative value indicates that the glycation is lower under the same average blood glucose. The glycated deviation value is used to characterize the difference in glycated tendency of individuals: in the onset evaluation model, it is used as an enhanced item for onset condition audit to identify people with higher risk under the same average blood glucose; in the control evaluation model, it is used as a trigger item for control condition audit to identify cases with abnormal glycated tendency despite long-term compliance.

[0072] The glycosylation accumulation model takes blood glucose time series data and glycated hemoglobin data as input, sets the high blood glucose threshold value to 7.8 mmol / L (the general numerical threshold value of postprandial hyperglycemia) according to the diabetes diagnosis standard, adopts a discrete implementable calculation process to obtain the threshold exposure integral, and then linearly synthesizes it with the glycated level to form a unified quantity that can be directly used for evaluation.

[0073] The blood glucose time series data is arranged in chronological order, and the blood glucose value in the i-th blood glucose time series data is represented by , and the time stamp is Protein glycation is a concentration-dependent non-enzymatic reaction, and blood glucose higher than the physiological safety upper limit will promote the formation of glycation products at a higher rate, so only the part exceeding the threshold value is calculated for accumulation, which can avoid equating fluctuations within the normal range to damage contribution.

[0074] Specifically, taking 7.8 mmol / L as the threshold value, taking the non-negative part exceeding the threshold value, using trapezoidal weighting on the amplitudes of the adjacent two points exceeding the threshold value, multiplying by the corresponding time difference, and accumulating to obtain the super-threshold exposure integral:

[0075] ;

[0076] Wherein, E represents the super-threshold exposure integral; N represents the total number of blood glucose time series data; i represents the i-th data in the blood glucose time series data; max represents the maximum value; represents the blood glucose value in the i-th blood glucose time series data; represents the blood glucose value in the i+1-th blood glucose time series data; represents the time difference between and .

[0077] The long-term accumulation of damage depends on the product of the non-negative part exceeding the threshold value and time, so the non-negative part exceeding the threshold value is time-weighted and integrated to obtain E, and then the average super-threshold exposure intensity is obtained by normalizing the total time to obtain the average super-threshold exposure intensity, which is used to describe the average intensity of continuous hyperglycemia during the statistical period. Linear synthesis of the average super-threshold exposure intensity and the glycosylated hemoglobin data obtains the glycosylation accumulation:

[0078] ;

[0079] Wherein, A represents the glycosylation accumulation; a represents the contribution weight of the glycosylated hemoglobin part; H represents the glycosylated hemoglobin data; β represents the contribution weight of the hyperglycemia exposure part; represents the average super-threshold exposure intensity.

[0080] It should be noted that, in order to ensure the dimensional consistency and reasonable contribution ratio of the glycosylated hemoglobin data and the average super-threshold exposure intensity, the coefficients a and β are calibrated offline in the model establishment stage. Specifically, a user sample data set with complete blood glucose time series data and glycosylated hemoglobin detection records is collected, the average super-threshold exposure intensity and the glycosylated hemoglobin data of each sample are obtained, and the late-stage advanced glycation end product level (corresponding to the glycosylation accumulation) is obtained. The least squares error between the model output glycosylation accumulation (calculated by the method of the present application) and the late-stage advanced glycation end product level is minimized, and the values of the coefficients a and β are obtained by linear regression. The obtained a and β are fixed constants and remain unchanged during the model running.

[0081] The glycated hemoglobin data reflects the average glycated level in 2-3 months, and the average super-threshold exposure intensity reflects the average intensity of high exposure during the statistical period, which respectively characterizes the cumulative mechanism from the background glycated and peak exposure; the linear synthesis of a and β forms a single glycosylation accumulation, which determines the long-term background and peak exposure in the same index.

[0082] The super-threshold exposure integral quantitatively represents the cumulative degree of threshold blood glucose during the statistical period; the average super-threshold exposure intensity provides a high exposure intensity representation aligned with unit time, which can distinguish the case of "average not high but peak frequent"; the glycosylation accumulation combines the long-term glycated level and the high exposure intensity into a single input, which is used in the control evaluation model to reflect the cumulative damage related risk, and in the incidence evaluation model to identify individual differences with high exposure characteristics when the average level is close.

[0083] The user data is processed by the life correction model, the glycated deviation model and the glycosylation accumulation model, and the analysis data including the corrected glycated hemoglobin data, the glycated deviation value and the glycosylation accumulation are output.

[0084] S3: When the population is determined as a non-diabetic population, an incidence evaluation model is established, the incidence evaluation model calculates a basic risk score based on the life correction model, modifies the basic risk score through incidence condition audit to obtain a glycosylation modified risk score, and maps the glycosylation modified risk score to an incidence risk grade.

[0085] Further, the corrected glycated hemoglobin data is obtained according to the life correction model, and the glycated deviation value is obtained according to the glycated deviation model.

[0086] The corrected glycated hemoglobin data is taken as the basic input of the incidence evaluation model, and the Sigmoid function is used to map the corrected glycated hemoglobin data to between 0 and 1 to obtain a basic risk score:

[0087] ;

[0088] The basic risk score represents the probability of risk occurrence, and the value range is [0, 1]. In order to avoid the boundary effect and non-linear imbalance of operation in the probability space, it is converted into the form of odds:

[0089] ;

[0090] Wherein, The basic risk score is represented by; The Sigmoid function is represented by k; the control slope is represented by; The corrected glycated hemoglobin data is represented by c; the center point of the risk curve is represented by; The probability of the basic risk score corresponds. The probability represents the relative ratio of event occurrence and non-occurrence, which is convenient for performing multiplicative or exponential correction.

[0091] It should be noted that the center point of the risk curve is set to 6.5% according to the international general diabetes diagnosis standard; reference epidemiological large cohort study (such as UKPDS, DCCT), the data of glycosylated hemoglobin and the risk of diabetes complications show a nonlinear relationship in the interval of 6.0%-8.0%, therefore, based on the user sample data set, by minimizing the difference between the basic risk score and the actual incidence rate, the control slope is obtained, so that the corrected glycosylated hemoglobin data is consistent with the rising trend of the actual incidence risk rate.

[0092] The basic risk score is corrected by the incidence condition audit:

[0093] According to the long-term follow-up epidemiological study (UKPDS, DCCT, Hempe, etc.), when the absolute value of the glycosylated deviation value is less than 0.5%, the deviation is mainly affected by the detection fluctuation and the short-term physiological factors; when the absolute value of the glycosylated deviation value is greater than or equal to 0.6%, it indicates that the individual glycosylation tendency is significantly higher than the average level of the group, which has independent physiological significance and is significantly related to the increase of the risk of diabetes complications. Therefore, the absolute value of the glycosylated deviation value 0.6% is used as the trigger condition of the incidence condition audit, if the absolute value of the glycosylated deviation value is greater than or equal to 0.6%, the glycosylated deviation value is called to correct the basic risk score, and the glycosylated corrected risk score is output, if the absolute value of the glycosylated deviation value is less than 0.6%, no correction is performed, and the glycosylated corrected risk score is equal to the basic risk score.

[0094] The glycosylated deviation value refers to the degree that the individual glycosylated hemoglobin data is higher or lower than the expected glycosylated hemoglobin data under the given average blood glucose, in the logistic regression, the glycosylated deviation value has a linear coefficient effect on the logarithmic probability, which is naturally consistent with the risk explanation. Specifically, the probability corresponding to the basic risk score is multiplied by the exponential function, the deviation value is taken as the input to calculate the risk amplification coefficient, and the dynamic adjustment of the risk score is realized. The correction formula is:

[0095] ;

[0096] Wherein, The probability corresponding to the glycosylated corrected risk score is represented by P; The probability corresponding to the basic risk score is represented by P; The exponential function is represented by exp; The gain coefficient of the glycosylated corrected risk score is represented by P; The glycosylated corrected risk score is obtained by:

[0097] ;

[0098] wherein, denotes the glycation correction risk score.

[0099] determined by regression analysis of glycation deviation value and diabetes incidence in user sample dataset, used to reflect the average amplification factor of incidence risk per 1% increase of glycation deviation.

[0100] The influence of glycation deviation value on risk is monotonous and continuous by exponential function correction on the odds form of the basic risk score. When the glycation deviation value is positive, the exponential term is greater than 1, and the risk score is adjusted upward. When the glycation deviation value is negative, the exponential term is less than 1, and the risk score remains or slightly decreases, thereby achieving automatic dynamic correction based on the degree of individual glycation deviation.

[0101] The glycation correction risk score is corrected according to the glycation accumulation, the intensity score is obtained by Sigmoid function compression of the glycation accumulation when any glycation correction condition is met, the gating coefficient is calculated according to the glycation correction risk score, the glycation correction risk score is calculated and output based on the gating coefficient and the intensity score, and if the glycation correction condition is not met, no correction is made, and the glycation correction risk score is equal to the glycation correction risk score.

[0102] The glycation correction condition includes two conditions:

[0103] The continuous duration of blood glucose time series data greater than 7.8 mmol / L is more than 30 minutes;

[0104] The cumulative duration of blood glucose time series data greater than 7.8 mmol / L in any 24 hours is more than 120 minutes.

[0105] During correction, the glycation accumulation is first compressed by Sigmoid function to convert the glycation accumulation into an intensity score in the interval of 0 to 1, to limit the excessive amplification of risk score by extreme high values. Subsequently, the gating coefficient is calculated according to the glycation correction risk score, which is equal to "1 minus the glycation correction risk score", used to reflect the risk up-regulation space. Finally, the gating coefficient and the intensity score are combined, and the glycation correction risk score is calculated and output by multiplicative correction with a pre-set gain coefficient. The correction formula is as follows:

[0106] ;

[0107] wherein, denotes the odds corresponding to the glycation correction risk score; denotes the gain coefficient of the glycation correction risk score; denotes the Sigmoid compression result of the glycation accumulation; denotes the glycation correction risk score.

[0108] convert the probability corresponding to the glycosylation correction risk score into the glycosylation correction risk score:

[0109]

[0110] wherein, represents the glycosylation correction risk score.

[0111] The threshold of 7.8 mmol / L in the glycosylation correction condition is derived from the international standard of postprandial hyperglycemia by the World Health Organization (WHO) and the American Diabetes Association (ADA); the time threshold of 30 minutes continuously and 120 minutes cumulatively in 24 hours is derived from the international blood glucose monitoring guidelines, reflecting the time length limit that blood glucose sustained threshold exposure may cause glycation reaction to accelerate. These two conditions ensure that the model can accurately capture the high exposure events that have actual impact on long-term glycation accumulation.

[0112] The glycosylation accumulation comprehensively reflects the long-term average glycation level and the intensity of hyperglycemic exposure, and has a time and amplitude double accumulation effect on glycation reaction. By Sigmoid compression, the inhibition of extreme values is realized, and by the dynamic adjustment of the gating coefficient, the correction amplitude is adjusted, so that the high exposure is more significant in the low baseline risk population, and remains stable in the high risk population, so as to ensure the stability of the evaluation while realizing the sensitive response of the risk assessment to the high exposure individuals.

[0113] Map the glycosylation correction risk score to the incidence risk grade, and divide the interval according to the preset threshold. Specifically, the glycosylation correction risk score interval is divided into three grades of low risk, medium risk and high risk. For example, when the glycosylation correction risk score is less than 0.4, it is determined as low risk, 0.4-0.7 is determined as medium risk, and higher than 0.7 is determined as high risk. Finally, according to the glycosylation correction risk score interval where the glycosylation correction risk score is located, the corresponding incidence risk grade is output.

[0114] S4: When the population is determined as a diabetic population, a control evaluation model is established, the control evaluation model calculates a basic control risk grade based on the life correction model and the glycosylation accumulation model, and the basic control risk grade is modified by control condition audit to obtain a control risk grade.

[0115] ​Further, the control evaluation model is based on a life correction model and a glycation accumulation model, and the glycated hemoglobin data and the glycation accumulation quantity jointly reflect the long-term control level and the cumulative effect of high blood glucose exposure of the diabetic patient. The life correction model is used to eliminate the measurement deviation caused by the difference in red blood cell life, so that the glycated hemoglobin data are comparable between different individuals; and the glycation accumulation model comprehensively reflects the cumulative effect of the duration and amplitude of high blood glucose exposure by combining the blood glucose time series data and the glycation level, thereby characterizing the long-term blood glucose control quality. By combining the two, the basic control risk level takes into account both the long-term average glycation level and the blood glucose fluctuation and exposure intensity.

[0116] Specifically, the corrected glycated hemoglobin data are obtained according to the life correction model, and the glycation accumulation quantity is obtained according to the glycation accumulation model.

[0117] The glycated hemoglobin data threshold and the glycation accumulation quantity threshold are set according to the diabetes diagnosis standard, the corrected glycated hemoglobin data are mapped to a long-term control level according to the glycated hemoglobin data threshold; and the glycation accumulation quantity is mapped to a cumulative exposure level according to the glycation accumulation quantity threshold, both the long-term control level and the cumulative exposure level are divided into three levels of low, medium and high, and the higher level of the long-term control level and the cumulative exposure level is taken as the basic control risk level.

[0118] The glycated hemoglobin data threshold is set according to the diabetes control standards of the World Health Organization (WHO) and the American Diabetes Association (ADA), and 7.0% is the main target value of blood glucose control. If the glycated hemoglobin data is less than 7.0%, it indicates good control, and the long-term control level is low risk; if the glycated hemoglobin data is between 7.0% and 8.0%, it indicates general control, and the long-term control level is medium risk; and if the glycated hemoglobin data is higher than 8.0%, it indicates poor control, and the long-term control level is high risk. The glycation accumulation quantity threshold is determined by statistical analysis of the user sample data set. According to the fitting relationship between the glycation accumulation quantity and the complication risk rate, the inflection point at which the complication risk rate significantly increases is taken as the division boundary to form three levels of low, medium and high.

[0119] The basic control risk level is corrected by control condition audit:

[0120] When the absolute value of the glycation deviation value is greater than or equal to 0.6%, the basic control risk level is adjusted according to the positive and negative directions of the glycation deviation value. If the glycation deviation value is positive, the basic control risk level is adjusted by one level, and the control risk level is obtained; if the glycation deviation value is negative, the basic control risk level remains unchanged, and the control risk level is equal to the basic control risk level.

[0121] If the absolute value of the glycation deviation value is less than 0.6%, no correction is performed, and the control risk level is equal to the basic control risk level.

[0122] The control condition audit correction mechanism is used for dynamically adjusting the control risk level when there is a glycation deviation. When the absolute value of the glycation deviation value is greater than or equal to 0.6%, it indicates that there is a difference between the measured glycosylated hemoglobin level and the average blood glucose that exceeds the normal fluctuation range. When the glycation deviation value is positive, it indicates that the patient has a higher glycation tendency under the same average blood glucose level, indicating that the long-term glycation process is more active, and the potential control risk is increased, so the basic control risk level is increased by one level; when the glycation deviation value is negative, it indicates that the individual glycation level is slightly lower than expected, which belongs to the physiological difference category and does not need to be adjusted.

[0123] The present application realizes the unified modeling of the first onset risk assessment of the non-diabetic population and the blood glucose control risk assessment of the diabetic population in the same method by constructing a risk assessment system based on the glycosylated hemoglobin data and the blood glucose time series data. The present application can automatically identify the population type according to the blood glucose and glycation characteristics of the user, and output the onset risk level and the control risk level respectively, thereby realizing the synchronous assessment of the health risks of the population at different stages, and providing a unified, objective and quantifiable assessment means for the early warning and later management of diabetes.

[0124] In an exemplary embodiment, the present application also provides a diabetes risk assessment system, comprising,

[0125] The acquisition module acquires user data, obtains a diabetes diagnosis standard, and performs population determination according to the user data and the diabetes diagnosis standard.

[0126] The calculation module establishes a life correction model, a glycation deviation model and a sugar glycation accumulation model based on the user data.

[0127] The onset assessment module establishes an onset assessment model when the population determination is a non-diabetic population, the onset assessment model calculates a basic risk score based on the life correction model, modifies the basic risk score through an onset condition audit to obtain a sugar glycation modified risk score, and maps the sugar glycation modified risk score to an onset risk level.

[0128] The control assessment module establishes a control assessment model when the population determination is a diabetic population, the control assessment model calculates a basic control risk level based on the life correction model and the sugar glycation accumulation model, and obtains a control risk level by modifying the basic control risk level through a control condition audit.

[0129] If the above functions are realized in the form of software function units and sold or used as independent products, they can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application essentially or parts of the present application that make contributions to the prior art or parts of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes.

[0130] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a list of executable instructions for implementing logic functions, which can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus or device, such as a computer-based system, a system including a processor or other system that can fetch instructions from an instruction execution system, apparatus or device and execute the instructions, or in conjunction with these instruction execution systems, apparatus or devices. For the purpose of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transport programs for use by or in connection with an instruction execution system, apparatus or device, or in conjunction with these instruction execution systems, apparatus or devices.

[0131] More specific examples (a non-exhaustive list) of the computer-readable medium include the following: an electrical connection having one or more wires (electrical devices), a portable computer diskette (magnetic devices), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CD ROM). In addition, the computer readable medium can even be paper or other suitable medium on which the program can be printed, because the program can be obtained electronically, for example, by optical scanning of the paper or other medium, followed by editing, interpreting or otherwise processing, if necessary, in other suitable ways, to be electronically obtained and then stored in the computer memory.

[0132] It should be understood that portions of the present application can be implemented in hardware, software, firmware, or combinations thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, implementation can be with any or a combination of the following technologies, which are all well known in the art: a discrete logic circuit having logic gates for implementing logic functions upon an application of data signals, an application specific integrated circuit having appropriate combinational logic gates, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0133] It should be noted that the above examples are only used to illustrate the technical solutions of the present application but not limit the present application. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced equivalently without departing from the spirit and scope of the present application, and all of them should be covered in the scope of the claims of the present application.

Claims

1. A method of assessing the risk of diabetes, characterized by, The method comprises the following steps: Collecting user data, obtaining diabetes diagnosis criteria, and performing population determination according to the user data and the diabetes diagnosis criteria; Establishing a life correction model, a glycation deviation model and a glycosylation accumulation model to process the user data and obtain analysis data; The establishment of the life correction model includes calculating the glycation rate coefficient according to the average blood glucose: ； wherein, represents a glycation rate coefficient; represents a constant; represents a glycation kinetics linear coefficient; represents the average blood glucose; the glycation rate coefficient is used to characterize the amount of glycation produced per unit of time at this average blood glucose level; Then, the equivalent red blood cell lifespan is calculated according to the glycosylated hemoglobin data and the glycation rate coefficient: ； wherein L represents the equivalent red blood cell life span; represents the measured glycated hemoglobin data; Finally, the glycosylated hemoglobin data is corrected based on the equivalent red blood cell lifespan, and the corrected glycosylated hemoglobin data is output: ； wherein, represents the corrected glycated hemoglobin data; represents the reference red blood cell lifespan; The establishment of the glycation deviation model includes calculating the expected glycosylated hemoglobin data according to the average blood glucose: ； wherein, represents expected glycated hemoglobin data; represents average blood glucose converted into units; subtracts expected glycated hemoglobin data from measured glycated hemoglobin data, and outputs a glycated deviation value; The establishment of the glycosylation accumulation model includes setting the high blood glucose threshold to 7.8 mmol / L according to the diabetes diagnosis criteria, and calculating the threshold exposure integral according to the blood glucose time series data: ； Wherein, E represents the super threshold exposure integral; N represents the total number of blood glucose time series data; i represents the i th data in the blood glucose time series data; max represents taking the maximum value; represents the blood glucose value in the i th blood glucose time series data; represents the blood glucose value in the i+1 th blood glucose time series data; represents and the time difference of The threshold exposure integral is normalized for the full time, the average threshold exposure intensity is calculated, and the average threshold exposure intensity and the glycosylated hemoglobin data are linearly synthesized to obtain the glycosylation accumulation; The analysis data includes the corrected glycosylated hemoglobin data, the glycation deviation value and the glycosylation accumulation; When the population determination is a non-diabetic population, a disease onset evaluation model is established, the disease onset evaluation model calculates a basic risk score based on the life correction model, the basic risk score is modified through disease onset condition audit to obtain a glycosylation modified risk score, and the glycosylation modified risk score is mapped to a disease onset risk level; The disease onset evaluation model calculates a basic risk score based on the life correction model, modifies the basic risk score through disease onset condition audit to obtain a glycosylation modified risk score, and maps the glycosylation modified risk score to a disease onset risk level, which includes obtaining the corrected glycosylated hemoglobin data according to the life correction model, and obtaining the glycation deviation value according to the glycation deviation model; The corrected glycosylated hemoglobin data is used as the basic input of the disease onset evaluation model, and the Sigmoid function is used to map the corrected glycosylated hemoglobin data to between 0 and 1 to obtain the basic risk score; The basic risk score is modified through disease onset condition audit: The basic risk score is modified according to the glycation deviation value, if the absolute value of the glycation deviation value is greater than or equal to 0.6%, the glycation deviation value is called, the basic risk score is modified through the exponential function, and the glycosylation modified risk score is output, if the absolute value of the glycation deviation value is less than 0.6%, no modification is performed, and the glycosylation modified risk score is equal to the basic risk score; The glycosylation modified risk score is modified according to the glycosylation accumulation, when any of the glycosylation modification conditions is met, the glycosylation accumulation is compressed by the Sigmoid function to obtain an intensity score, a gating coefficient is calculated according to the glycosylation modified risk score, and the glycosylation modified risk score is calculated and output based on the gating coefficient and the intensity score, if the glycosylation modification condition is not met, the glycosylation modified risk score is equal to the glycosylation modified risk score; The glycosylation modification conditions include two conditions: The continuous time length of the blood glucose time series data greater than 7.8 mmol / L is 30 minutes; The cumulative time length of the blood glucose time series data greater than 7.8 mmol / L within any 24 hours is 120 minutes; mapping the glycosylation modified risk score to a morbidity risk level; when the population is determined as a diabetes population, establishing a control evaluation model, the control evaluation model is based on a life correction model and a glycosylation accumulation model to calculate a basic control risk level, and the basic control risk level is modified by a control condition audit to obtain a control risk level; the control evaluation model is based on a life correction model and a glycosylation accumulation model to calculate a basic control risk level, and the basic control risk level is modified by a control condition audit to obtain a control risk level, including obtaining corrected glycosylated hemoglobin data according to the life correction model, and obtaining glycosylation accumulation according to the glycosylation accumulation model; setting glycosylated hemoglobin data threshold and glycosylation accumulation threshold according to diabetes diagnosis standard, mapping the corrected glycosylated hemoglobin data to long-term control level according to the glycosylated hemoglobin data threshold, and mapping the glycosylation accumulation to cumulative exposure level according to the glycosylation accumulation threshold, both the long-term control level and the cumulative exposure level are divided into three levels of low, medium and high, and the higher level of the long-term control level and the cumulative exposure level is taken as the basic control risk level; the basic control risk level is modified by a control condition audit: when the absolute value of glycosylation deviation value is greater than or equal to 0.6%, the basic control risk level is adjusted according to the positive and negative directions of the glycosylation deviation value, if the glycosylation deviation value is positive, the basic control risk level is adjusted by one level, and the control risk level is obtained; if the glycosylation deviation value is negative, the basic control risk level remains unchanged, and the control risk level is equal to the basic control risk level; if the absolute value of glycosylation deviation value is less than 0.6%, no modification is made, and the control risk level is equal to the basic control risk level.

2. A method of assessing the risk of diabetes according to claim 1, wherein: the user data includes glycosylated hemoglobin data, fasting blood glucose data and blood glucose time series data; the diabetes diagnosis standard is a numerical threshold for diagnosing diabetes, and the user data and the diabetes diagnosis standard are compared to perform population determination: when the fasting blood glucose data is greater than or equal to 7.0mmol / L or the glycosylated hemoglobin data is greater than or equal to 6.5%, the population is determined as a diabetes population; when the fasting blood glucose data is less than 7.0mmol / L and the glycosylated hemoglobin data is less than 6.5%, the population is determined as a non-diabetes population.

3. A diabetes risk assessment system for use in a diabetes risk assessment method according to any one of claims 1 to 2, characterized in that including, a collection module, collecting user data, obtaining diabetes diagnosis standard, and performing population determination according to user data and diabetes diagnosis standard; a calculation module, establishing a life correction model, a glycosylation deviation model and a glycosylation accumulation model, processing user data to obtain analysis data; a morbidity evaluation module, when the population is determined as a non-diabetes population, establishing a morbidity evaluation model, the morbidity evaluation model is based on a life correction model to calculate a basic risk score, the basic risk score is modified by a morbidity condition audit to obtain a glycosylation modified risk score, and the glycosylation modified risk score is mapped to a morbidity risk level; The control evaluation module, when the crowd is determined as a diabetic crowd, establishes a control evaluation model, the control evaluation model calculates a basic control risk level based on a life correction model and a glycosylation accumulation model, and the basic control risk level is corrected through control condition auditing to obtain a control risk level.

Images