Glycosylated hemoglobin value correction method, model construction device and system, computer equipment and storage medium

Abstract

The invention discloses a glycosylated hemoglobin value correction method, a model construction device, a model construction system, computer equipment and a storage medium, and relates to the technical field of glycosylated hemoglobin detection. Respectively carrying out regression analysis by using measured values of large sample patients according to erythrocyte lifetime zoning conditions; and correspondingly fitting a corrected glycosylated hemoglobin value. According to the method, patients are subjected to grouping modeling according to red blood cell lifetime intervals (smaller than or equal to 65 days, 66-89 days and larger than or equal to 90 days), independent regression equations are adopted in different intervals, the problem that a single formula is poor in applicability in different red blood cell lifetime intervals is solved, more accurate HbA1c data are provided for doctors due to the use of a correction model, and the accuracy of the HbA1c data is improved. The method can assist in evaluating the blood glucose control level of a patient and avoid wrong treatment decisions caused by underestimation or overestimation of HbA1c, so that more accurate blood glucose reducing drug adjustment and treatment scheme design are realized.

Description

Technical Field

[0001] The present invention relates to the technical field of glycated hemoglobin detection, and in particular to a glycated hemoglobin value correction method, a model building device, a system, a computer device and a storage medium. Background Art

[0002] Blood sugar management is the core task of diabetes prevention and control; whether accurate blood sugar management can be truly achieved is the key to determining whether diabetic patients can obtain long-term benefits from cardiovascular, renal and other complications; and the accuracy of glycosylated hemoglobin test values ​​is a prerequisite for achieving accurate blood sugar management; glycosylated hemoglobin is currently the "gold standard" for clinical diagnosis and adjustment of hypoglycemic treatment, but the accuracy of its test value is affected to a great extent by the life span of red blood cells in addition to the blood sugar level; therefore, the glycosylated hemoglobin level of diabetic patients whose red blood cell life span deviates from the normal range cannot truly reflect the patient's blood sugar level; such patients urgently need to eliminate the influence of red blood cell life span on glycosylated hemoglobin and obtain their true glycosylated hemoglobin level; however, there is currently no technical means in clinical practice at home and abroad to adjust the glycosylated hemoglobin test value for red blood cell life span, which leads to the glycosylated hemoglobin test value of diabetic patients with shortened red blood cell life span being lower than their actual blood sugar level for a long time, and ultimately these patients cannot receive appropriate hypoglycemic treatment; therefore, how to correct the influence of red blood cell life span on glycosylated hemoglobin is a major issue that needs to be solved urgently in clinical practice.

[0003] However, for half a century, the principle of the traditional determination method of RBCs life span is to use radioactive substances or non-radioactive markers to label certain components of RBCs, and obtain RBCs according to the attenuation or metabolic clearance rate of the markers, such as 51Cr labeling method, 15N-glycine labeling method, biotin labeling method, etc. The above traditional determination methods can intuitively and accurately determine the survival time of RBCs, but have many disadvantages such as long time consumption, complicated operation, high cost, repeated blood sampling and adverse reactions, which greatly limit the clinical application and promotion of the above detection methods; due to the limitations of conditions such as RBC life span detection equipment, At present, there is still a lack of technical means to study the impact of RBCs variation on HbA1C test values ​​in a large sample of type 2 diabetes population, and how to correct the impact of red RBCs on glycated hemoglobin test values. The team of Professor Ma Yongjian of Shenzhen University invented a new safe and convenient red blood cell detection technology based on the principle of CO exhalation test, which provides technical conditions for realizing large-sample red blood cell life span detection. This project intends to conduct a multi-center, large-sample diabetic population study based on my country's independently developed red blood cell life span detection technology and machine autonomous learning technology, and establish a correction system and verification system for correcting the effect of red blood cell life span on glycated hemoglobin test values. Summary of the invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problems solved by the present invention are: the existing glycated hemoglobin detection method has the problems of insufficient correction accuracy due to the abnormal red blood cell life span, poor applicability of the single correction formula, and an imperfect model verification system; and how to construct a correction model suitable for different red blood cell life span ranges to achieve accurate correction of the HbA1c detection value.

[0006] To solve the above technical problems, the present invention provides the following technical solutions: a method for correcting glycosylated hemoglobin values, comprising zoning the life span of red blood cells; using the measured values ​​of a large sample of patients to perform regression analysis according to the red blood cell life span zoning conditions; and fitting the corrected glycosylated hemoglobin values ​​accordingly.

[0007] As a preferred embodiment of the method for correcting the glycated hemoglobin value of the present invention, the partitioning includes partitioning the red blood cell life span based on the biasing effect of the red blood cell life span on the glycated hemoglobin measurement value.

[0008] As a preferred embodiment of the method for correcting the glycated hemoglobin value of the present invention, the measured values ​​include the glycated hemoglobin measured value and the life span of red blood cells.

[0009] As a preferred solution of the glycated hemoglobin value correction method of the present invention, wherein: the partition includes a partition threshold value set to a node with obvious bias influence.

[0010] As a preferred embodiment of the method for correcting the glycated hemoglobin value of the present invention, the regression analysis includes performing restricted spline regression analysis of the glycated hemoglobin variation index and the red blood cell life span, finding nodes with obvious bias effects, and constructing a correction model.

[0011] As a preferred embodiment of the method for correcting the glycated hemoglobin value of the present invention, the construction of the correction model includes, when the life span of red blood cells is ≤ 65 days, the correction model is expressed as:

[0012] HbA1c'=-0.00459×RBCs+0.5919×HbA1c+3.289

[0013] Wherein, HbA1C' represents the corrected glycated hemoglobin value, RBCs represents the red blood cell life span, and HbA1c represents the glycated hemoglobin measurement value.

[0014] As a preferred embodiment of the method for correcting the glycated hemoglobin value of the present invention, the construction of the correction model further includes that when 66 days ≤ red blood cell life span ≤ 89 days, the correction model is expressed as:

[0015] HbA1c'=0.0000435×RBCs+0.5687×HbA1+3.097

[0016] The corrected glycated hemoglobin value is output according to the red blood cell life span.

[0017] A model building device for glycated hemoglobin value correction includes a grouping module for data preprocessing and interval division of red blood cell life span; a regression analysis module for fitting correction values ​​through multivariate linear regression and fitting regression equations for each interval; and a model verification module for model verification based on a modeling cohort, an internal verification cohort, and an external verification cohort.

[0018] As a preferred solution of the model building device for glycated hemoglobin value correction of the present invention, the grouping module is specifically used for the interval division of the red blood cell life span, including dividing the collected large sample patients into a modeling cohort, an internal verification cohort and an external verification cohort, and calculating the overall regression equation of the linear correlation between HbA1c and GA based on the HbA1c and GA of the modeling cohort and the internal verification cohort, which is expressed as:

[0019] eHbA1c=0.217×GA+3.466

[0020] Where GA represents the glycated albumin value, and eHbA1c represents the estimated glycated hemoglobin value. The glycated hemoglobin variability index HGI is calculated as follows:

[0021] HGI=HbA1c-eHbA1c

[0022] Among them, HbA1c represents the measurement value of glycated hemoglobin. The restricted spline regression model of HGI and red blood cell life span has an inflection point at 65 days, and HGI is significantly negatively correlated with red blood cell life span. Patients are divided into three subgroups according to red blood cell life span ≤ 65 days, 66 days ≤ red blood cell life span ≤ 89 days, and red blood cell life span ≥ 90 days.

[0023] As a preferred solution of the model building device for glycated hemoglobin value correction of the present invention, the regression analysis module is specifically used to fit the regression equations between the partitions, including the bias effect of red blood cell life on the glycated hemoglobin measurement value, the glycated hemoglobin value and the red blood cell life as independent variables, and the glycated hemoglobin correction model is constructed by a multivariate linear regression equation; when the red blood cell life is ≤65 days, the correction model substituted is expressed as:

[0024] HbA1c'=-0.00459×RBCs+0.5919×HbA1c+3.289

[0025] Among them, HbA1C' represents the corrected glycated hemoglobin value, RBCs represents the life span of red blood cells, and HbA1c represents the glycated hemoglobin measurement value; when 66 days ≤ red blood cell life span ≤ 89 days, the correction model is expressed as:

[0026] HbA1c'=0.0000435×RBCs+0.5687×HbA1+3.097

[0027] The corrected glycated hemoglobin value is output according to the red blood cell life span.

[0028] As a preferred solution of the model building device for glycated hemoglobin value correction described in the present invention, the model verification module is specifically used for the model verification, including using the modeling cohort and the internal verification cohort to cross-validate the reliability of the model, testing the applicability of the formula on new data through the external verification cohort, and verifying the ability of the corrected glycated hemoglobin value to distinguish high sugar state based on the C index and evaluation indicators.

[0029] Another object of the present invention is to provide a glycated hemoglobin value correction system, which can calculate the corrected glycated hemoglobin value by establishing a correction model based on red blood cell life span partitions, thereby solving the problem of poor applicability of current glycated hemoglobin detection technology.

[0030] As a preferred solution of the glycated hemoglobin value correction system described in the present invention, it includes: a data processing module, a correction model construction module, and a blood sugar management module; the data processing module includes a red blood cell life acquisition module, a glycated hemoglobin value acquisition module, and a glycated albumin value acquisition module, and the data acquisition module is used for the red blood cell life acquisition module, the glycated hemoglobin value acquisition module, and the glycated albumin value acquisition module to collect the patient's red blood cell life value, glycated hemoglobin value, and glycated albumin value accordingly, and optimize the patient preparation process and the gas sampling and detection process; the correction model construction module is used to establish a correction model based on red blood cell life partitions and calculate the corrected glycated hemoglobin value; the blood sugar management module is used to obtain the real blood sugar level in the population with shortened red blood cell life based on the corrected glycated hemoglobin value.

[0031] A computer device includes a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of a method for correcting a glycated hemoglobin value.

[0032] A computer-readable storage medium stores a computer program, which implements the steps of a method for correcting a glycated hemoglobin value when executed by a processor.

[0033] Beneficial effects of the present invention: The glycated hemoglobin value correction method provided by the present invention introduces the red blood cell life span as a correction variable, constructs a multivariate regression correction model between partitions, and solves the problem of deviation of HbA1c detection value due to abnormal red blood cell life span; in particular, for patients with significantly shortened or prolonged red blood cell life span (such as anemic patients, blood transfusion patients), the accuracy of glycated hemoglobin value can be significantly improved; by grouping patients according to the red blood cell life span interval (≤65 days, 66-89 days, ≥90 days) for modeling, independent regression equations are used for different intervals, thus overcoming the problem of poor applicability of a single formula in different red blood cell life span intervals; the model is applicable to a wide range of people with a red blood cell life span of 30-120 days; through a large sample A multicenter study (3023 patients) was conducted to construct a modeling cohort and independent internal and external validation cohorts, and the C index and other indicators were used to verify the model performance to ensure the stability and reliability of the correction model. The verification showed that the corrected HbA1c value significantly improved the ability to distinguish high glucose status in all cohorts. The use of the correction model provided doctors with more accurate HbA1c data, which can assist in evaluating the patient's blood glucose control level and avoid erroneous treatment decisions caused by underestimation or overestimation of HbA1c, thereby achieving more accurate hypoglycemic drug adjustment and treatment plan design. The present invention has achieved better results in terms of HbA1c detection accuracy, optimizing model applicability, providing verification data support, and assisting personalized diagnosis and treatment. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0035] Figure 1 This is an overall flow chart of a method for correcting glycosylated hemoglobin values ​​provided in the first embodiment of the present invention.

[0036] Figure 2 A logic flow chart of a method for correcting glycosylated hemoglobin value provided in the third embodiment of the present invention.

[0037] Figure 3 A graph showing the relationship between the levels of glycated albumin (%) and glycated hemoglobin (%) and the life span of red blood cells in a modeling cohort of a glycated hemoglobin value correction method provided in the third embodiment of the present invention.

[0038] Figure 4 This is a Spearman correlation analysis diagram of red blood cell life span and hemoglobin variation glycation index in type 2 diabetic patients according to a method for correcting glycosylated hemoglobin values ​​provided in the third embodiment of the present invention.

[0039] Figure 5 A relationship diagram between the red blood cell life span of type 2 diabetic patients and the HGI restricted spline regression fitting curve according to a method for correcting glycosylated hemoglobin values ​​provided in the third embodiment of the present invention.

[0040] Figure 6 A comparison chart of glycated hemoglobin, glycated albumin, glycated hemoglobin estimation value and hemoglobin variation glycation index for different red blood cell life span groups according to a glycated hemoglobin value correction method provided in the third embodiment of the present invention.

[0041] Figure 7 A diagnostic effect and calibration diagram of an evaluation correction formula for a method for correcting glycosylated hemoglobin values ​​provided in the third embodiment of the present invention.

[0042] Figure 8 This is an overall flow chart of a glycated hemoglobin value correction system provided in the fourth embodiment of the present invention. DETAILED DESCRIPTION

[0043] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0044] Example 1, reference Figure 1 , as an embodiment of the present invention, provides a method for correcting a glycated hemoglobin value, comprising:

[0045] S1: Partitioning of red blood cell life span.

[0046] Furthermore, a specific step of S1 includes finding significant nodes using restricted spline regression based on the biased effect of red blood cell lifespan on glycated hemoglobin measurement values, and dividing red blood cell lifespan into partitions of ≤65 days, 66-89 days, and ≥90 days.

[0047] It should be noted that the partitioning includes partitioning the red blood cell life span based on the biasing effect of the red blood cell life span on the glycated hemoglobin measurement.

[0048] It should also be noted that red blood cell life span can be calculated by a non-invasive CO breath test. The non-invasive CO breath test method for red blood cell life span provides a convenient and safe source of input data for the calibration model, avoiding the high cost, high complexity and potential adverse reactions of traditional labeling methods for detecting red blood cell life span. Glycated hemoglobin measurement values ​​can be determined by venous blood sampling.

[0049] S2: Use the measurements of a large sample of patients to perform regression analysis based on the red blood cell life span divisions.

[0050] Furthermore, a specific step of S2 includes using a large sample data of no less than 1000 patients, taking the glycated hemoglobin measurement value and the red blood cell life span as independent variables and the covariate as auxiliary based on the partition, and constructing a partition correction model.

[0051] It should be noted that the measured values ​​include glycated hemoglobin measurement and red blood cell life span.

[0052] It should also be noted that other optional variables include patient gender, age, glycated albumin (GA), etc., to evaluate the impact of potential covariates on the model; when establishing the correction model, adding optional variables (such as patient gender, age, glycated albumin GA, etc.) as covariates to evaluate their potential impact on the bias of HbA1c measurement values ​​will help optimize the fitting effect and universality of the model; the introduction of multiple variables enables the correction model to adapt to the individual characteristics of different patients and improve its applicability to complex populations (such as elderly patients or patients with other metabolic abnormalities).

[0053] It should also be noted that the partitioning includes nodes where the partitioning threshold is set to nodes where the bias effect is significant.

[0054] It should also be noted that the regression analysis included restricted spline regression analysis of the glycosylated hemoglobin variability index and red blood cell life span to find nodes with obvious bias effects and construct a correction model.

[0055] It should also be noted that a preferred scheme for partitioning includes dividing patients into three groups according to the red blood cell life span range based on the distribution characteristics of RBC life span: red blood cell life span ≤ 65 days, 66 days ≤ red blood cell life span ≤ 89 days, and red blood cell life span ≥ 90 days. When the red blood cell life span is ≥ 90 days, it is considered a normal red blood cell life span, so there is no need to establish a corresponding correction model.

[0056] It should also be noted that the construction of the correction model includes when the red blood cell life span is ≤ 65 days, the correction model is expressed as:

[0057] HbA1c'=-0.00459×RBCs+0.5919×HbA1c+3.289

[0058] Wherein, HbA1C' represents the corrected glycated hemoglobin value, RBCs represents the red blood cell life span, and HbA1c represents the glycated hemoglobin measurement value.

[0059] It should also be noted that the construction of the correction model also includes when 66 days ≤ red blood cell life span ≤ 89 days, the correction model is expressed as:

[0060] HbA1c'=0.0000435×RBCs+0.5687×HbA1c+3.097

[0061] The corrected glycated hemoglobin value is output according to the red blood cell life span.

[0062] It should also be noted that based on the biased effect of red blood cell lifespan on glycated hemoglobin measurement values, zoning and constructing different linear regression models can more accurately reflect the differentiated effects of red blood cell lifespan on HbA1c measurement values ​​in different intervals; the zoning thresholds (65th day, 90th day) are obtained through inflection point analysis of red blood cell lifespan and HbA1c variability index, ensuring the scientificity and rationality of the interval division; zoning correction solves the problem that a single formula is difficult to take into account all red blood cell lifespan intervals, thereby improving the accuracy of the correction value.

[0063] It should also be noted that traditional methods mostly use HbA1c as the only variable for linear regression correction, ignoring the key role of red blood cell lifespan in the formation of HbA1c; by using HbA1c and red blood cell lifespan as dual independent variables, the established model can accurately capture the multi-dimensional correlation between HbA1c and RBC lifespan, improving the accuracy and applicability of the correction model; red blood cell lifespan significantly affects the generation rate of HbA1c and the deviation of its final detection value. Taking RBC lifespan as the core independent variable, this influence can be directly quantified in the correction model, making the correction result closer to the patient's actual blood glucose level; combining HbA1c and red blood cell lifespan can achieve accurate correction based on individual patient characteristics, avoiding the problem of inaccurate correction of traditional models for specific patients (such as patients with anemia, hemolysis, etc.), and directly improving the reliability of diagnosis and treatment plans.

[0064] S3: Correspondingly fit the corrected glycated hemoglobin value.

[0065] Furthermore, a specific step of S3 includes substituting the red blood cell life span and the glycated hemoglobin measurement value according to the partition correction model to calculate the corrected glycated hemoglobin value to more truly reflect the patient's long-term blood glucose level.

[0066] It should be noted that by designing a special correction formula for patients with red blood cell lifespan ≤ 65 days, the deviation of HbA1c test value caused by significantly shortened red blood cell lifespan can be corrected more accurately; especially for patients with hemolytic diseases, anemia or other patients with abnormally short red blood cell lifespan, the correction model ensures that the test value is closer to the actual blood glucose level and optimizes the scope of application of the model; for patients with red blood cell lifespan between 66 and 89 days, a different correction model is used to eliminate the impact of slight deviation of RBC lifespan from the normal range on HbA1c test value; this model can accurately correct the test value error caused by changes in red blood cell metabolic rate, for patients Provide more reliable blood glucose level assessment data; the correction model for the medium red blood cell lifespan group fills the gap in traditional correction methods and solves the problem of poor performance of simple linear formulas in this interval; by designing correction formulas for different intervals, the corresponding model is dynamically selected according to the range of red blood cell lifespan, avoiding the problem of insufficient applicability of a single formula; this partitioned correction method can be compatible with more patient groups, from low red blood cell lifespan (≤65 days) to near-normal lifespan (66-89 days), and can provide accurate correction values; the method of dynamically adjusting the correction formula according to RBC lifespan improves the universality of the model and meets the individualized diagnosis and treatment needs of different patients in clinical practice.

[0067] Example 2 is an embodiment of the present invention, which provides a model building device for glycated hemoglobin value correction, including: a grouping module 100, a regression analysis module 200, and a model verification module 300; the grouping module 100 is used for data preprocessing and interval division of red blood cell life span; the regression analysis module 200 is used to fit the correction value through multivariate linear regression, and the regression equation is fitted for each interval respectively; the model verification module 300 is used to perform model verification based on the modeling cohort, the internal verification cohort, and the external verification cohort.

[0068] Furthermore, the grouping module 100 is specifically used for the interval division of the red blood cell life span, including dividing the collected large sample patients into a modeling cohort, an internal verification cohort, and an external verification cohort, and calculating the overall regression equation of the linear correlation between HbA1c and GA based on the HbA1c and GA of the modeling cohort and the internal verification cohort, which is expressed as:

[0069] eHbA1c=0.217×GA+3.466

[0070] Where GA represents the glycated albumin value, and eHbA1c represents the estimated glycated hemoglobin value. The glycated hemoglobin variability index HGI is calculated as follows:

[0071] HGI=HbA1c-eHbA1c

[0072] Among them, HbA1c represents the measurement value of glycated hemoglobin. The restricted spline regression model of HGI and red blood cell life span has an inflection point at 65 days, and HGI is significantly negatively correlated with red blood cell life span. Patients are divided into three subgroups according to red blood cell life span ≤ 65 days, 66 days ≤ red blood cell life span ≤ 89 days, and red blood cell life span ≥ 90 days.

[0073] It should be noted that the regression analysis module 200 is specifically used to fit the regression equations for the intervals, including the bias effect of the red blood cell life span on the glycated hemoglobin measurement value, the glycated hemoglobin value and the red blood cell life span are independent variables, and the glycated hemoglobin correction model is constructed through the multivariate linear regression equation; when the red blood cell life span is ≤65 days, the correction model inserted is expressed as:

[0074] HbA1c'=-0.00459×RBCs+0.5919×HbA1c+3.289

[0075] Among them, HbA1C' represents the corrected glycated hemoglobin value, RBCs represents the life span of red blood cells, and HbA1c represents the glycated hemoglobin measurement value; when 66 days ≤ red blood cell life span ≤ 89 days, the correction model is expressed as:

[0076] HbA1c'=0.0000435×RBCs+0.5687×HbA1c+3.097

[0077] The corrected glycated hemoglobin value is output according to the red blood cell life span.

[0078] It should also be noted that the model verification module 300 is specifically used to perform model verification, including using the modeling cohort and the internal verification cohort to cross-validate the reliability of the model, testing the applicability of the formula on new data through the external verification cohort, and verifying the ability of the corrected glycated hemoglobin value to distinguish high sugar states based on the C index and evaluation indicators.

[0079] Example 3, reference Figure 2-Figure 7 , which is an embodiment of the present invention, provides a method for correcting glycosylated hemoglobin values. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0080] Figure 2It is a logical flow chart, which shows the operation steps of each link. The patient preparation stage: gas collection after getting up in the early morning and before 12 noon. The patient should be in a fasting state before gas collection, and gas collection should be avoided after strenuous exercise (gas collection operation should be performed in a resting state); the subject has not received red blood cell transfusion within one week before gas collection, and cannot actively or passively smoke within 24 hours; the gas collection operation is performed according to standard steps, and the patient is required to provide the hemoglobin value of the day; the physiological, pathological conditions and medication status of the subjects before the test are consulted and understood, so as to facilitate a comprehensive judgment on the reasons for the changes in the life span of red blood cells; note: gastrointestinal bleeding, strenuous exercise, staying up late, infection, etc. can cause a transient increase in red blood cell destruction, that is, a transient shortening of the life span of red blood cells; taking certain drugs will make the red blood cell destruction rate unstable in the short term; gas collection and blood collection operations are performed on patients who meet the gas collection requirements.

[0081] Gas sampling and testing process: Assemble the alveolar gas collection device, connect the blowing nozzle, alveolar air bag and cavity air bag through a tee; connect the blowing nozzle to the rear end of the tee, the cavity air bag to the front end of the tee, and the alveolar air bag to the side end of the tee; ask the patient to take a deep breath and hold his breath for about 10 seconds; then exhale all the chest gas at once through the blowing nozzle; check whether the air bag is full (press the air bag with your fingers to see if it is concave by more than 1 cm). If not, squeeze the air in the air bag with your hands to empty it, and repeat the above blowing steps; seal the alveolar air bag for testing; at the same time, use an air bag to collect the background gas in the environment.

[0082] S4: Red blood cell life span detection: Place the air bag and the patient's alveolar bag into the corresponding channels respectively, input the patient's hemoglobin value for the day, and the patient's red blood cell life span can be obtained.

[0083] S5: Glycated hemoglobin and glycated albumin testing: The glycated hemoglobin and glycated albumin values ​​of each patient are obtained by venous blood sampling and sending the blood to the same laboratory for testing.

[0084] S6: Establishment of the correction model: Referring to Table 1, first, 3023 patients with type 2 diabetes from different communities of the Tianjin Diabetic Retinopathy Screening Cohort were collected and randomly divided into a modeling cohort (n=2409) and an internal validation cohort (n=614) in a ratio of 4:1; 477 patients with type 2 diabetes admitted to the "Three One Care" clinic of the Zhu Xianyi Memorial Hospital of Tianjin Medical University were selected as the external validation cohort.

[0085] Table 1 General clinical characteristics of the modeling cohort, internal validation cohort, and external validation cohort

[0086]

[0087] refer to Figure 3The HbA1c and GA of 3023 patients were used to calculate the overall regression equation for the linear correlation between HbA1c and glycated albumin (GA), that is, the estimated value of glycated hemoglobin eHbA1c = 0.217 × GA + 3.466, the model goodness of fit R = 0.775, and the result was significant p < 0.001; the glycated hemoglobin variability index (HGI) is the difference between the HbA1c test value and eHbA1c.

[0088] S7: Reference Figure 4-Figure 5 , since the restricted spline regression model of HGI and RBC lifespan had an inflection point at 65 days, the correlation index r = 0.075 indicated that there was a weak negative correlation between red blood cell lifespan and glycated albumin variability index, and the result significance p = 0.0002 showed that this correlation was statistically significant, that is, this negative correlation was not a random or accidental phenomenon, so HGI was significantly negatively correlated with RBC lifespan; the patients were divided into three subgroups according to RBC lifespan ≤ 65 days, 66 days ≤ RBC lifespan ≤ 89 days, and RBC lifespan ≥ 90 days.

[0089] refer to Figure 6 , compared the differences in glycated indicators such as HbA1c, GA, estimated HbA1c (eHbA1c), and HGI in patients with different RBC lifespan groups (RBC lifespan ≤65 days, 66-89 days, and ≥90 days); Figure 6 The AC graph data in the figure show that the HbA1c, GA, and eHbA1c values ​​of the RBC lifespan ≤ 65 days group were significantly lower than those of the other groups, especially when compared with the RBC lifespan ≥ 90 days group, the significant difference was more obvious (marked with "*" in the figure); this indicates that short red blood cell lifespan can lead to an underestimation of the patient's glycation indexes; Figure 6 Figure D shows that the HGI (Glycated Hemoglobin Variability Index) has a wider fluctuation range in the RBC lifespan ≤65 days group, indicating that the error of HbA1c test values ​​in patients with low RBC lifespan is more significant; the data support the necessity of correcting the bias of HbA1c measurement due to low RBC lifespan through a partition correction model; when the RBC lifespan is less than 90 days, the actual blood glucose level of this population is underestimated to a certain extent, especially in the RBC lifespan ≤65 days group.

[0090] S8: Based on the different influencing rules of RBC lifespan and the linear relationship between HbA1c and eHbA1c, the glycated hemoglobin correction formula was constructed by a multivariate linear regression equation with eHbA1c as the dependent variable and HbA1c and RBC lifespan as independent variables; RBC lifespan ≤ 65 days group: HbA1c' = -0.00459 × RBCs + 0.5919 × HbA1c + 3.289, model goodness of fit R = 0.6690; 66 days ≤ RBC lifespan ≤ 89 days group: HbA1c' = 0.0000435 × RBCs + 0.5687 × HbA1C + 3.097, model goodness of fit R = 0.7606.

[0091] refer to Figure 7 The corrected HbA1C showed good discrimination of the hyperglycemic state of all cohorts: the C index of the modeling cohort was 0.8515 (95% CI 0.8296-0.8735, indicating that the fluctuation range of the model's discrimination ability was small and the performance was stable). The C index of the internal validation and external validation cohorts was similar to that of the modeling cohort, which were 0.8504 (95% CI 0.8052-0.8957) and 0.8884 (95% CI 0.8488-0.9280, indicating that the external validation effect was relatively stable and better than the modeling cohort), respectively. CI represents the confidence interval. The C index is an indicator used to evaluate the discriminatory ability of the prediction model, and is especially commonly used in the performance evaluation of the binary classification model in clinical medicine. Referring to Table 2, the clinical characteristics of patients with different red blood cell lifespans in the modeling cohort were recorded.

[0092] Table 2 Clinical characteristics of patients with different red blood cell lifespans in the modeling cohort

[0093]

[0094]

[0095] Obtaining the glycated hemoglobin correction value: For patients with shortened red blood cell lifespan, the glycated hemoglobin correction value can be calculated based on their red blood cell lifespan by substituting the measured glycated hemoglobin and red blood cell lifespan values ​​into different correction formulas.

[0096] Example 4, reference Figure 8 , which is an embodiment of the present invention, provides a glycated hemoglobin value correction system, including a data processing module 400, a correction model building module 500, and a blood glucose management module 600.

[0097] The data processing module 400 includes a red blood cell lifespan acquisition module 401, a glycosylated hemoglobin value acquisition module 402, and a glycosylated albumin value acquisition module 403. The data acquisition module 400 is used for the red blood cell lifespan acquisition module 401, the glycosylated hemoglobin value acquisition module 402, and the glycosylated albumin value acquisition module 403 to collect the patient's red blood cell lifespan value, glycosylated hemoglobin value, and glycosylated albumin value, and optimize the patient preparation process and the gas sampling detection process;

[0098] Furthermore, S9: in the initial stage of system operation, the data processing module 400 is started, which is mainly responsible for collecting basic data related to the patient and optimizing the data collection process; the red blood cell lifespan collection module 401 quickly obtains the patient's red blood cell lifespan value through a non-invasive CO exhalation test, and transmits the result to the system for recording and storage; the glycosylated hemoglobin value collection module 402 measures the patient's glycosylated hemoglobin (HbA1c) value through venous blood sampling, and inputs it into the correction model as raw data; the glycosylated albumin value collection module 403 detects the patient's glycosylated albumin (GA) value for evaluating short-term blood glucose levels, and at the same time provides auxiliary variables for the correction model; during the data collection process, the system optimizes the patient preparation process and gas sampling detection operations to ensure the accuracy and consistency of data collection, and provide reliable input data for subsequent correction models.

[0099] The correction model building module 500 is used to establish a correction model based on the red blood cell life span partition and calculate the corrected glycosylated hemoglobin value;

[0100] It should be noted that, S10: when data collection is completed, the system automatically switches to the correction model construction module 500, and corrects the glycated hemoglobin value by the partition modeling method; based on the partition logic of red blood cell life (such as ≤65 days, 66-89 days, ≥90 days), the patients are divided into different correction groups; the patient data of each partition (red blood cell life, glycated hemoglobin value, glycated albumin value) are subjected to multivariate regression analysis to calculate the corrected glycated hemoglobin value (cHbA1c); for example: for patients with red blood cell life ≤65 days, the system uses a specific regression formula to correct HbA1c; for patients with red blood cell life ≥90 days, a normal grouping model is used or no correction is required; the correction model automatically outputs and stores the corrected glycated hemoglobin value to provide a basis for subsequent analysis.

[0101] The blood glucose management module 600 is used to obtain the real blood glucose level in the population with shortened red blood cell life span based on the corrected glycosylated hemoglobin value.

[0102] Furthermore, S11: after the correction model outputs the results, the blood glucose management module 600 evaluates the patient's blood glucose management status according to the corrected glycated hemoglobin value; in the patient group with shortened red blood cell lifespan, the corrected glycated hemoglobin value (cHbA1c) more truly reflects the patient's long-term blood glucose level, and solves the problem of underestimation of HbA1c due to abnormal red blood cell lifespan; the system analyzes the correction results to determine whether the patient has poor blood glucose control (such as cHbA1c>7%), and generates corresponding blood glucose management recommendations; for patients with poor blood glucose control, the system prompts the doctor to adjust the treatment plan (such as optimizing drug dosage, adjusting diet plan, etc.); for patients with good blood glucose control, the system provides follow-up monitoring recommendations to help patients maintain a good state.

[0103] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0104] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0105] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering, or processing in another suitable manner as necessary, and then stored in a computer memory.

[0106] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc. It should be noted that the above embodiments are only used to illustrate the technical solution of the present invention and are not limited. Although the present invention is described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention, which should be included in the scope of the claims of the present invention.

[0107] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention 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 invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for correcting glycosylated hemoglobin, characterized in that: include: Zoning of red blood cell life span; Using the measurements of a large sample of patients, regression analysis was performed separately according to the red blood cell life span partition; The corrected glycated hemoglobin value was fitted accordingly.

2. The method for correcting glycosylated hemoglobin value according to claim 1, characterized in that: The partition includes: Red cell life span was partitioned based on its biasing effect on glycated hemoglobin measurements.

3. The method for correcting glycosylated hemoglobin value according to claim 1 or 2, characterized in that: The measurements include, Glycated hemoglobin measurements and red blood cell life span.

4. The method for correcting glycosylated hemoglobin value according to claim 1 or 2, characterized in that: The partitioning includes nodes where the partitioning threshold is set to have a significant bias effect.

5. The method for correcting glycosylated hemoglobin value according to any one of claims 1, 3 or 4, characterized in that: The regression analysis includes performing restricted spline regression analysis on the glycosylated hemoglobin variability index and the red blood cell life span, finding nodes with obvious bias effects, and constructing a correction model.

6. The method for correcting glycosylated hemoglobin value according to claim 5, characterized in that: The constructing of the correction model comprises: When the life span of red blood cells is ≤65 days, the correction model is expressed as: HbA1c'=-0.00459×RBCs+0.5919×HbA1c+3.289 Wherein, HbA1C' represents the corrected glycated hemoglobin value, RBCs represents the red blood cell life span, and HbA1c represents the glycated hemoglobin measurement value.

7. The method for correcting glycosylated hemoglobin value according to claim 5, characterized in that: The constructing of the correction model also includes: When 66 days ≤ red blood cell life span ≤ 89 days, the correction model is expressed as: HbA1c'=0.0000435×RBCs+0.5687×HbA1+3.097 The corrected glycated hemoglobin value is output according to the red blood cell life span.

8. A model building device for calibrating glycosylated hemoglobin values, characterized in that: include, A grouping module (100), used for data preprocessing and interval division of red blood cell life span; A regression analysis module (200) is used to fit the correction value by multivariate linear regression, and to fit the regression equation for each interval; The model verification module (300) is used to perform model verification based on a modeling queue, an internal verification queue and an external verification queue.

9. The model building device for glycated hemoglobin value correction according to claim 8, characterized in that: The grouping module (100) is specifically used for the interval division of the red blood cell life span, comprising dividing the collected large sample patients into a modeling cohort, an internal verification cohort, and an external verification cohort, and calculating the overall regression equation of the linear correlation between HbA1c and GA based on the HbA1c and GA of the modeling cohort and the internal verification cohort, which is expressed as: eHbA1c=0.217×GA+3.466 Among them, GA represents glycated albumin value, and eHbA1c represents the estimated glycated hemoglobin value; The glycosylated hemoglobin variability index (HGI) was calculated as: HGI=HbA1c-eHbA1c Among them, HbA1c represents the measurement value of glycated hemoglobin. The restricted spline regression model of HGI and red blood cell life span has an inflection point at 65 days, and HGI is significantly negatively correlated with red blood cell life span. Patients are divided into three subgroups according to red blood cell life span ≤ 65 days, 66 days ≤ red blood cell life span ≤ 89 days, and red blood cell life span ≥ 90 days.

10. The model building device for glycated hemoglobin value correction according to claim 8, characterized in that: The regression analysis module (200) is specifically used to fit regression equations between the partitions respectively, including the biased influence of red blood cell life span on the glycated hemoglobin measurement value, the glycated hemoglobin value and the red blood cell life span are independent variables, and a glycated hemoglobin correction model is constructed through a multivariate linear regression equation; When the life span of red blood cells is ≤65 days, the correction model is expressed as: HbA1c'=-0.00459×RBCs+0.5919×HbA1c+3.289 Wherein, HbA1C' represents the corrected glycated hemoglobin value, RBCs represents the red blood cell life span, and HbA1C represents the glycated hemoglobin measurement value; When 66 days ≤ red blood cell life span ≤ 89 days, the correction model is expressed as: HbA1c'=0.0000435×RBCs+0.5687×HbA1+3.097 The corrected glycated hemoglobin value is output according to the red blood cell life span.

11. The model building device for glycated hemoglobin value correction according to claim 7, characterized in that: The model validation module (300) is specifically used for the model validation, including using a modeling cohort and an internal validation cohort to cross-validate the reliability of the model, testing the applicability of the formula on new data through an external validation cohort, and validating the ability of the corrected glycated hemoglobin value to distinguish between high glucose states based on the C index and evaluation indicators.

12. A system for calibrating glycosylated hemoglobin values ​​using any one of claims 1 to 6, characterized in that: It includes a data processing module (400), a correction model building module (500), and a blood sugar management module (600); The data processing module (400) comprises a red blood cell lifespan acquisition module (401), a glycosylated hemoglobin value acquisition module (402), and a glycosylated albumin value acquisition module (403). The data acquisition module (400) is used for the red blood cell lifespan acquisition module (401), the glycosylated hemoglobin value acquisition module (402), and the glycosylated albumin value acquisition module (403) to collect the red blood cell lifespan value, glycosylated hemoglobin value, and glycosylated albumin value of the patient, thereby optimizing the patient preparation process and the gas sampling detection process. The correction model building module (500) is used to establish a correction model based on the red blood cell life span partition and calculate the corrected glycated hemoglobin value; The blood glucose management module (600) is used to obtain the real blood glucose level in a population with shortened red blood cell lifespan based on the corrected glycosylated hemoglobin value.

13. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method for correcting a glycated hemoglobin value according to any one of claims 1 to 6 are implemented.

14. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for correcting a glycated hemoglobin value according to any one of claims 1 to 7 are implemented.

Images