Multi-dimensional dynamic continuous biological reference interval generation method and related equipment

Abstract

The invention provides a multi-dimensional dynamic continuous biological reference interval generation method and related equipment, and is applied to the technical field of data processing. The method comprises the steps of obtaining target user data; processing the target user data to generate to-be-monitored user data with identification information; a total training set matched with the identification information and a preset multivariate continuous biological reference interval model matched with the identification information are obtained, and the total training set comprises a training sample corresponding to a to-be-detected target item of the target user; preprocessing the total training set to generate a target training set and a target test set; training a preset multivariate continuous biological reference interval model based on the target training set and the target test set to generate a target multivariate continuous biological reference interval model; the to-be-monitored user data is processed based on the target multivariate continuous biological reference interval model, a prediction result is generated, and the prediction result comprises the detection state of the to-be-detected disease of the target user.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a method for generating a multi-dimensional dynamic continuous biological reference interval and related equipment. Background Art

[0002] Establishing appropriate biological reference intervals for test items, referred to as reference intervals (commonly known as normal ranges), is crucial to ensuring that clinicians correctly diagnose and treat diseases. Whether the test results are "normal" directly depends on the appropriateness of the reference intervals. Establishing accurate reference intervals is an important basis for clinical judgment of whether various diseases exist and whether the treatment is effective. Currently, clinical laboratories use a unified reference interval, or a reference interval established by zoning based on age or gender. Most of these reference intervals come from the manufacturer's reagent instructions or current industry standards.

[0003] However, the reference intervals widely used in clinical laboratories currently have limitations, which are mainly reflected in the following aspects: Limitations of unified reference intervals: Many laboratories use unified reference intervals provided by manufacturers or based on industry standards, which do not take into account individual differences. Insufficiency of the zoning method: Although the zoning method takes into account the influence of factors such as age or gender, it still cannot dynamically reflect the gradual changes of test indicators with continuous variables such as age. Ignoring individual differences: Test indicators such as thyroid hormones will change with non-continuous variables such as gender, region, and continuous variables such as age, season, and BMI. The unified reference interval cannot reflect these individual differences. Discontinuity of physiological changes: The reference interval based on the zoning method may not accurately reflect normal physiological changes near the boundaries of the age segments. For example, according to the current domestic reference interval industry standards, the upper reference limit of creatinine is significantly different between 59-year-old and 60-year-old men. This "jump" obviously cannot reflect normal physiological fluctuations.

[0004] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present disclosure, and therefore may include information that does not constitute the prior art known to ordinary technicians in the field. Summary of the invention

[0005] The purpose of this application is to provide a method for generating a multi-dimensional dynamic continuous biological reference interval and related equipment and systems, which at least to a certain extent overcome the problems existing in the prior art, and establish a model of continuous reference intervals that can reflect multiple continuous variables by simultaneously considering non-continuous variables and continuous variables, aiming to capture the dynamic changes of test indicators under normal physiological conditions, improve the accuracy of clinical tests, reduce misdiagnosis and missed diagnosis, provide support for personalized medicine, and enable clinicians to manage and treat diseases more effectively.

[0006] Other features and advantages of the present application will become apparent from the following detailed description, or may be learned in part by the practice of the present invention.

[0007] According to one aspect of the present application, a method for generating a multidimensional dynamic continuous biological reference interval and related equipment are provided, including: acquiring target user data; processing the target user data to generate user data to be monitored with identification information, wherein the identification information is used to characterize the target item to be detected by the target user at present; acquiring a total training set matching the identification information and a preset multivariate continuous biological reference interval model matching the identification information, wherein the total training set includes training samples corresponding to the target items to be detected by the target user; preprocessing the total training set to generate a target training set and a target test set; training the preset multivariate continuous biological reference interval model based on the target training set and the target test set to generate a target multivariate continuous biological reference interval model; processing the user data to be monitored based on the target multivariate continuous biological reference interval model to generate a prediction result, wherein the prediction result includes the detection status of the disease to be detected by the target user.

[0008] In one embodiment of the present application, the preprocessing of the total training set to generate a target training set and a target test set includes: extracting features from the total training set to determine an original feature library, wherein the original feature library includes non-continuous variable data and continuous variable data; processing the total training set based on the non-continuous variable data to generate an initial training set; processing the initial training set based on the continuous variable data to generate a target training set, wherein the target training set includes several biological reference intervals, and different biological reference intervals correspond to different disease types.

[0009] In one embodiment of the present application, the preset multivariate continuous biological reference interval model is trained based on the target training set and the target test set to generate a target multivariate continuous biological reference interval model, including: processing the target training set to generate multiple continuous variable data matching the identification information; processing the multiple continuous variable data based on a preset function to generate a target biological reference interval, wherein the target biological reference interval is a preset reference range matching the target item; and generating a target multivariate continuous biological reference interval model based on the target biological reference interval.

[0010] In one embodiment of the present application, the processing of the target user data to generate user data to be monitored with identification information includes: processing the target user data to generate user attribute information, wherein the user attribute information includes target items to be detected by the target user; performing feature extraction on the target user data to generate a feature data set; processing the feature data set based on the original feature library to generate patient feature information; processing the patient feature information based on the user attribute information to generate user data to be monitored with identification information.

[0011] In one embodiment of the present application, the user data to be monitored is processed based on the target multivariate continuous biological reference interval model to generate a prediction result, including: if the patient characteristic information represents that the target user is a patient of a target hospital, then obtaining the case information of the target user; based on the case information of the target user, obtaining the historical laboratory indicator data of the target user; based on the target multivariate continuous biological reference interval model, the historical laboratory indicator data is processed to generate a historical prediction result.

[0012] In one embodiment of the present application, the data of the user to be monitored is processed based on the target multivariate continuous biological reference interval model to generate a prediction result, and also includes: obtaining real-time laboratory indicator data of the user to be monitored; processing the laboratory indicator data based on the target multivariate continuous biological reference interval model to generate real-time vital sign information of the user to be monitored; processing the real-time vital sign information of the user to be monitored based on the historical prediction results to generate the detection results of the target items of the user to be monitored within a preset time period.

[0013] In one embodiment of the present application, the total training set is preprocessed to generate a target training set and a target test set, and the method further includes: processing the target training set to generate an initial candidate group and an initial fixed group; processing the initial candidate group and the initial fixed group respectively to generate a preset candidate group and a preset fixed group, wherein the preset candidate group and the preset fixed group respectively carry identification information representing different covariate priority information; generating a loop iteration rule based on the identification information, wherein the loop iteration rule is set based on the number of covariates to be matched preset by the user; generating a loop iteration rule based on the loop iteration rule ... The covariates in the preset candidate group and the covariates in the preset fixed group are processed according to the generation rules to generate a preprocessed candidate group and a preprocessed fixed group, wherein the covariates include continuous variables and categorical variables; the preprocessed candidate group and the preprocessed fixed group are processed based on the preset matching rules to generate a matching result; the matching result is processed based on the variable numerical calculation model to generate the statistical value of the preprocessed candidate group and the statistical value of the preprocessed fixed group; if the statistical value of the preprocessed candidate group and the statistical value of the preprocessed fixed group meet the matching requirements respectively, a target candidate group and a target fixed group are generated.

[0014] Another aspect of the present application is a device for generating a multi-dimensional dynamic continuous biological reference interval, characterized in that it includes: an acquisition module, used to acquire target user data; acquire a total training set matching the identification information and a preset multivariate continuous biological reference interval model matching the identification information, wherein the total training set includes training samples corresponding to the target items to be detected for the target user; a processing module, used to process the target user data to generate user data to be monitored with identification information, wherein the identification information is used to characterize the target items to be detected by the target user; pre-process the total training set to generate a target training set and a target test set; train the preset multivariate continuous biological reference interval model based on the target training set and the target test set to generate a target multivariate continuous biological reference interval model; process the user data to be monitored based on the target multivariate continuous biological reference interval model to generate a prediction result, wherein the prediction result includes the detection status of the disease to be detected by the target user.

[0015] According to another aspect of the present application, an electronic device is characterized in that it includes: a first processor; and a memory for storing executable instructions of the first processor; wherein the first processor is configured to execute the above-mentioned method for generating a multi-dimensional dynamic continuous biological reference interval by executing the executable instructions.

[0016] According to another aspect of the present application, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a second processor, the method for generating the multi-dimensional dynamic continuous biological reference interval is implemented.

[0017] According to another aspect of the present application, a computer program product is provided, including a computer program, characterized in that when the computer program is executed by a third processor, the computer program implements the above-mentioned method for generating multi-dimensional dynamic continuous biological reference intervals.

[0018] The present application provides a method for generating a multi-dimensional dynamic continuous biological reference interval and related equipment, wherein the server obtains target user data; processes the target user data to generate user data to be monitored with identification information, wherein the identification information is used to characterize the target item to be detected by the target user; obtains a total training set matching the identification information and a preset multivariate continuous biological reference interval model matching the identification information, wherein the total training set includes training samples corresponding to the target item to be detected by the target user; pre-processes the total training set to generate a target training set and a target test set; trains the preset multivariate continuous biological reference interval model based on the target training set and the target test set to generate a target multivariate continuous biological reference interval model; processes the user data to be monitored based on the target multivariate continuous biological reference interval model to generate a prediction result, wherein the prediction result includes the detection status of the target user's disease to be detected. By simultaneously considering non-continuous variables and continuous variables, a model of continuous reference intervals that can reflect multiple continuous variables is established, aiming to capture the dynamic changes of test indicators under normal physiological conditions, improve the accuracy of clinical tests, reduce misdiagnosis and missed diagnosis, provide support for personalized medicine, and enable clinicians to manage and treat diseases more effectively.

[0019] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 A flow chart showing a method for generating a multi-dimensional dynamic continuous biological reference interval provided by an embodiment of the present application;

[0021] Figure 2 A schematic structural diagram of a device for generating a multi-dimensional dynamic continuous biological reference interval provided by an embodiment of the present application is shown;

[0022] Figure 3 A schematic diagram of the structure of an electronic device provided by an embodiment of the present application is shown;

[0023] Figure 4 A schematic diagram of a storage medium provided by an embodiment of the present application is shown;

[0024] Figure 5 A 3D diagram showing the continuous reference intervals of a test item provided by an embodiment of the present application as it changes with age and month is shown;

[0025] Figure 6 A schematic diagram of fluctuations of a fitting curve provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0026] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0027] Combine the following Figure 1 To describe the method for generating a multi-dimensional dynamic continuous biological reference interval according to an exemplary embodiment of the present application. It should be noted that the following application scenarios are only shown to facilitate understanding of the spirit and principles of the present application, and the embodiments of the present application are not limited in this regard. On the contrary, the embodiments of the present application can be applied to any applicable scenario.

[0028] Thyroid hormones play a vital role in human metabolism and growth and development. The main thyroid hormones include thyroxine (T4), triiodothyronine (T3), free T3 (FT3), (FT4) and thyroid stimulating hormone (TSH), commonly known as the five thyroid hormones. They are of great significance for the clinical diagnosis of various thyroid-related diseases, such as hyperthyroidism, hypothyroidism, and various thyroiditis. For example, according to the current health industry standards in my country, the reference ranges for the Roche detection system are as follows: T3, 1.3-2.4nmol / L; T4, 70-140nmol / L; FT3, 3.85-6.30pmol / L; FT4, 12.80-21.30pmol / L; TSH, 0.75-5.6mIU / L. However, many studies have shown that the test indicators in the human body will change dynamically with non-continuous variables such as gender and region, as well as continuous variables such as age and season (month). For example, thyroid-related hormone levels vary with age and seasons. In this case, a unified thyroid-related hormone reference interval cannot dynamically reflect the physiological changes caused by age and seasonal changes. Some reference intervals based on age zoning methods, although taking into account the differences in reference intervals caused by factors such as age, cannot reflect the gradual dynamic changes caused by multiple continuous variables. Especially for people near the boundaries of the reference interval, there will be sudden changes in the reference interval, which is inconsistent with the slow physiological changes in reality and may lead to incorrect health assessments or even misdiagnosis of diseases.

[0029] In one embodiment, the present application also proposes a method for generating a multi-dimensional dynamic continuous biological reference interval and related equipment. Figure 1 The flowchart of a method for generating a multi-dimensional dynamic continuous biological reference interval according to an embodiment of the present application is schematically shown. Figure 1 As shown, the method is applied to a server, comprising:

[0030] S101, obtaining target user data.

[0031] In one embodiment, the user's basic information, including gender, age, location, etc., and health status, including past medical history, current symptoms, lifestyle, etc., are collected. The user's weight and height are recorded to calculate the BMI (body mass index); the current test time is recorded to consider the impact of season or month on the test results.

[0032] S102: Process the target user data to generate user data to be monitored with identification information.

[0033] In one implementation, the target user data is processed to generate user attribute information, wherein the user attribute information includes a target item to be detected by the target user; features are extracted from the target user data to generate a feature data set; the feature data set is processed based on an original feature library to generate patient feature information; the patient feature information is processed based on the user attribute information to generate user data to be monitored with identification information, wherein the identification information is used to characterize the target item to be detected by the target user.

[0034] First, determine the target items that the user needs to detect, collect the user's basic information, including but not limited to name, age, gender, region, etc., and create attribute information containing the user's target items to be detected based on the collected data. Extract key features from the user data, such as physiological indicators, lifestyle, etc., and organize the extracted features into data sets for analysis and model training.

[0035] The feature data set is preprocessed by standardization and normalization, and the feature data set is processed to generate detailed patient feature information, and the user attribute information is combined with the patient feature information to ensure the relevance of the data. In addition, the unique identifier of the test item, user ID, test time and other information are added to the data set to ensure that the generated data set is accurate and reflects the user status.

[0036] In another implementation, the data is cleaned and preprocessed to remove outliers, fill in missing values, and perform necessary transformations; basic statistical indicators such as mean, median, standard deviation, etc. are calculated to describe the central tendency and variability of the data; the correlation between different variables is analyzed, such as the relationship between age and BMI, the relationship between season and specific test indicators, etc.; the data is stratified according to non-continuous variables such as gender, age, and region to analyze the characteristics of each subgroup separately; time factors are considered to analyze the changing trends of test indicators over time. Multivariate statistical methods, such as multivariate regression analysis, are used to explore the comprehensive impact of multiple variables on test results.

[0037] S103, obtaining a total training set matching the identification information and a preset multivariate continuous biological reference interval model matching the identification information.

[0038] In one embodiment, the total training set includes training samples corresponding to the target items to be tested for the target user. Retrieve the corresponding identification information from the database, select the training samples related to the identification information, form the total training set, clean the training samples, remove invalid or erroneous data, perform standardization, select features related to the test items from the training set, such as gender, age, etc., determine the statistical or machine learning model suitable for analyzing multivariate continuous variables, use the total training set data to train the preset reference interval model, evaluate the accuracy and generalization ability of the model through cross-validation and other methods, adjust the model according to the verification results, improve its prediction performance, apply the trained model to new data, and predict individualized biological reference intervals.

[0039] S104, preprocessing the total training set to generate a target training set and a target test set.

[0040] In one embodiment, data of one or some test items of all physical examination populations in a physical examination center of a certain unit within 5 years are collected, and the above data are preliminarily screened for inclusion in the group using multi-dimensional information in the hospital HIS system, such as blood pressure, BMI, ultrasound examination, CT examination, laboratory test indicators such as liver function, renal function, serum index (indices of sample quality, such as hemolysis index, lipemia index, jaundice index), etc. The specific screening criteria are related to the clinical significance and influencing factors of the target project. Exclude repeated individual results. If there are 3 test results of the same person (same medical record number) within 5 years, only the most recent result is retained. If there are population regions or selectivity, other types of populations are excluded. For example, if a continuous reference interval for a certain project of the Chinese population is established, other national populations must be excluded. The collected data are tested for the data distribution of the test results of the target project using statistical software such as R, SPSS, Python, SAS and other software. According to whether the data distribution is normal or whether it can be converted to normal, select the corresponding outlier elimination method (such as Tukey method, Z score method) to eliminate the data after the outliers for further analysis.

[0041] In another embodiment, feature extraction is performed on the total training set to determine an original feature library, wherein the original feature library includes non-continuous variable data and continuous variable data; the total training set is processed based on the non-continuous variable data to generate an initial training set; the initial training set is processed based on the continuous variable data to generate a target training set, wherein the target training set includes several biological reference intervals, and different biological reference intervals correspond to different disease types.

[0042] Identify and extract key features from the total training set, including non-continuous variables (such as gender, region) and continuous variables (such as age, BMI, season), create an original feature library containing all extracted features for subsequent analysis and model training, and use the non-continuous variable data in the original feature library to classify and group the total training set to generate an initial training set. The initial training set is a data set that has been processed with non-continuous variables, providing a basis for subsequent continuous variable processing. The continuous variable data in the original feature library are used to further analyze the initial training set, such as trend analysis, correlation analysis, etc. Based on the results of continuous variable processing, a target training set is generated from the initial training set. This data set will be used to establish a biological reference interval model.

[0043] In another embodiment, the data of the preliminary screening are compared between groups according to the classification of non-continuous variables using software such as R, SPSS, Python, and SAS to determine the reference interval grouping. The data are classified according to different non-continuous variables (such as region, gender). After classification, the corresponding test method is selected to compare the results of the purpose test items between groups. Such as t-test, one-way analysis of variance, Mann-Whitney U test, Kruskal-Wallis H test. For groups with significant statistical differences, continuous reference interval tests are performed after grouping according to this factor. For the above statistics, P<0.05 is considered to be statistically different.

[0044] In another embodiment, the data is classified according to continuous variable factors that may affect the level of the target item (such as age, month, BMI). The classification can be in the form of segmentation (such as referring to domestic industry standards, adults can be grouped by ten years); the months are grouped according to spring (March, April, May), summer (June, July, August), autumn (September, October, November), and winter (December, January, February); BMI can determine the grouping boundaries according to specific test items and clinical significance. After grouping, the test results of the target items are compared between groups.

[0045] According to the grouping situation (two or more groups) and data distribution (normal or non-normal), select appropriate statistical methods for inter-group comparison (such as one-way analysis of variance, Mann-Whitney U test, t test, Kruskal-Wallis H test). For groups with significant statistical differences, the grouping factors should be used as continuous reference interval variables to establish continuous reference intervals. In addition, correlation analysis should be performed on different continuous variable factors that may affect the level of the target item and the test results of the target item (such as Pearson correlation analysis, Spearman correlation analysis, etc.). If the factor is significantly correlated with the test result data, the variable should also be used as one of the influencing factors of the continuous reference interval for subsequent modeling. For the above statistics, P<0.05 is considered to be statistically different.

[0046] After determining the grouping method based on non-continuous variables and the continuous variable elements that are significantly related to the test items, start to model the continuous reference interval for each subgroup. For example, if according to the above scheme, it is determined that only gender and age are effective influencing factors for a certain test item, then the applicant must first divide the data into male and female groups based on gender, and then model the age-related continuous reference interval for each subgroup separately. If it is determined that region, age and month are effective influencing factors for a certain test item, then the applicant must first group the data based on region, and then model the continuous reference interval of the two continuous variables of age and month for each subgroup separately.

[0047] S105, training the preset multivariate continuous biological reference interval model based on the target training set and the target test set to generate a target multivariate continuous biological reference interval model.

[0048] In one embodiment, the target training set is processed to generate multiple continuous variable data that match the identification information; based on a preset function, such as a regression function or a smoothing function (including but not limited to a spline function), the multiple continuous variable data are processed to generate a target biological reference interval, wherein the target biological reference interval is a preset reference range that matches the target item; and a target multivariate continuous biological reference interval model is generated based on the target biological reference interval.

[0049] Specifically, the preset function is y=β0+β1x+β2x 2 +β3x 3 +…+β n x n +∈, where y is the dependent variable (target project outcome), x is the independent variable (such as age, month, BMI), β0,β1,β2,…,β n is the regression coefficient, which needs to be estimated by fitting the data, ∈ is the error term, and n is the order of the polynomial.

[0050] In addition, the preset function can also be Q y (τ|x)=β0+β1x1+β2x2+…+β n x n ; where Q y (τ|x) is the τth quantile of y given x, where y is the dependent variable (target item outcome) and x is the independent variable (such as age, month, BMI). τ is generally defined as 2.5, 97.5 or 95.

[0051] In addition, the preset function can also be S i (x) = a i +b i (xx i )+c i (xx i ) 2 +d i (xx i ) 3 ; Given a set of data points (x1,y1),(x2,y2),…,(x m ,y n ), where y is the dependent variable (target program outcome) and x is the independent variable (e.g., age, month, BMI).

[0052] The goal of cubic spline interpolation is to find a set of piecewise cubic polynomials S i (x), where the formula satisfies the following conditions:

[0053] 1. S i (x i )=y i For all i;

[0054] 2. S i (x i+1 )=y i+1 For all i;

[0055] 3. S' i (x i+1 )=S' i+1 (x i+1 ) for all i;

[0056] 4. S″ i (x i+1 )=S″ i+1 (x i+1 ) for all i.

[0057] First, the target training set is processed to extract multiple continuous variable data matching the user identification information, and these data are smoothed using a preset regression function or smoothing function to capture the nonlinear relationship in the data. Based on the data processed by the spline function, biological reference intervals matching the target items are generated. These intervals represent the expected fluctuation range under normal physiological conditions. These data are used to construct a target multivariate continuous biological reference interval model, which can reflect the biological indicators of different disease types. The generalized additive mixed model (GAMLSS) is used to model dual continuous variables, which is a flexible statistical method that can handle data with complex distribution characteristics.

[0058] Select appropriate spline functions and adjust model parameters to ensure that the model can capture the key features of the data while avoiding overfitting or underfitting. By evaluating residual plots, histograms, and QQ plots, check whether the distribution of residuals conforms to normality, thereby evaluating the stability and goodness of fit of the model. Use cross-validation and other methods to verify the model to ensure that it has good predictive and generalization capabilities. Based on the evaluation results, further optimize and adjust the model to improve its performance. Apply the final model to actual data to generate accurate biological reference intervals.

[0059] In the target training set, several biological reference intervals are determined according to biological indicators and disease types to ensure that different biological reference intervals correspond to corresponding disease types for clinical application. The target training set is optimized, including data balancing and feature engineering, to improve the effect of model training. The target training set is prepared for the training of multivariate continuous biological reference interval models to ensure that the model can learn the characteristics of different disease types. After model training, the performance of the model on the target training set is evaluated to ensure its accuracy and reliability. The trained model is applied to actual clinical data, and feedback is collected for iterative optimization of the model.

[0060] In another embodiment, the present application establishes data 97.5 and 2.5 percentile curves (surfaces) (bilateral reference intervals, test indicators are meaningful when they are high or low) or 95 percentile curves (surfaces) (unilateral reference intervals, the reference lower limit is 0, and the indicators are meaningful only when they are high) to represent the reference intervals that conform to normal physiological fluctuations by adopting curve (surface) fitting or regression. The modeling method is preliminarily selected according to the data distribution characteristics. The models that can be selected are polynomial regression, Altman method, quantile regression, dual-core method, Kosmic method, LMS (lambda-median-sigma) method, generalized additive model, generalized linear model, multivariate linear model, random forest, GAMLSS method. Each method has its own characteristics, and can be evaluated and selected according to the distribution characteristics of the data, as well as using Akaike information criterion (AIC) and Bayesian information criterion (BIC). The smaller the AIC and BIC, the better the performance of the model in terms of balancing complexity and goodness of fit, and the stronger the predictive ability. In practical applications, the applicant tends to use the GAMLSS method to construct the reference interval. The GAMLSS method extends the generalized linear model and generalized additive model, and can model multiple aspects of the data, including location (mean), scale (standard deviation), shape (skewness and kurtosis). There are more than 100 parameters in the model that can be used to model the response variable, which makes the modeling flexible and controllable. Taking thyroid stimulating hormone (TSH) as an example, the GAMLSS model is used to establish a continuous reference interval for different gender groups with age. The specific steps are as follows: 1. The cleaned TSH data are grouped according to gender, and the TSH data of each gender group are randomly divided into a training set (80 percent) and a validation set (20 percent). 2. Using R's "gamlss pakage", a model is established based on the TSH training set data, and the response variable "TSH result" is explained by the independent variable "age" (univariate continuous reference interval model), or the response variable "TSH result" is explained by the independent variables "age" and "month" (multivariate continuous reference interval model). 3. First, in GAMLSS, select the appropriate distribution family according to the distribution characteristics of TSH data. If the data is normally distributed, select "Normal Distribution (NO)" when modeling. If it is right-skewed data, select "Gamma Distribution (GA)". The real TSH data is skewed. In order to eliminate the influence of possible outliers in the data and ensure the robustness of the model, the distribution family selected for this project is "BOX-COX-t". This distribution can be used to transform the data into a normal state before modeling. GAMLSS supports a variety of spline functions for non-parametric smoothing and modeling of complex data patterns.

[0061] Establish optional splines for continuous reference intervals. Commonly used splines include: B-spline, P-spline, natural spline, cubic spline, and penalized spline. In this project, all the above spline functions were used to establish multiple fitting curve regression models for TSH results changing with age, and the AIC and BIC results in each model were compared. Through comparison, the cubic spline with the smallest AIC and BIC was selected to perform continuous reference interval modeling and fit the 97.5 and 2.5 percentile curves (surfaces) in this model. Observe the fluctuations of the TSH fitting curve (surface), and use professional experience to determine whether it conforms to normal physiological fluctuations to avoid underfitting or overfitting. If there are abnormal physiological fluctuations, adjust the degrees of freedom to convert it into a better fit. As follows Figure 6 As shown in Figure A, the overall level trend of the test items decreases with age, but there are too many small fluctuations in the fitting line, which obviously does not conform to normal physiological changes and is overfitting. The curve in Figure B is smoother and more in line with the normal fluctuation trend of the indicator with age, which is obviously better.

[0062] In another embodiment, the multivariate continuous reference interval model established in the present application can be directly embedded in the software system of the detection instrument, or the intermediate system of the detection instrument. When the above system software is connected to the hospital laboratory information system (LIS), it can receive information such as the patient's age, gender, and sample collection time, and then calculate and transmit the specific reference interval of the corresponding individual into the LIS system, which is reflected in the patient's test report. In addition, this model can also be directly embedded in the LIS system to directly reflect the specific reference interval of the corresponding individual, which is reflected in the patient's test report, achieving more personalized health assessment and disease diagnosis.

[0063] S106, processing the user data to be monitored based on the target multivariate continuous biological reference interval model to generate a prediction result.

[0064] In one implementation, the prediction result includes the detection status of the disease to be detected by the target user.

[0065] Acquire real-time laboratory indicator data of the user to be monitored; process the laboratory indicator data based on the target multivariate continuous biological reference interval model to generate real-time vital sign information of the user to be monitored; process the real-time vital sign information of the user to be monitored based on historical prediction results to generate test results of target items of the user to be monitored within a preset time period.

[0066] The target multivariate continuous biological reference interval model is applied to the data of the user to be monitored to predict the detection status of the disease and obtain the real-time laboratory indicator data of the user to be monitored. These data reflect the user's current health status. The real-time laboratory indicator data are processed using the target multivariate continuous biological reference interval model to generate the user's real-time vital sign information. The real-time vital sign information is further analyzed in combination with the user's historical prediction results. Based on the historical data and real-time vital sign information, a personalized reference interval for the target item and the final "normal" or "abnormal" judgment result are generated within a preset time period, and the final positive or negative result is judged according to the reference interval.

[0067] Integrate real-time vital sign information with historical prediction results to form a comprehensive user health profile, evaluate prediction results, determine whether the user's test status is normal or abnormal, and whether further medical intervention is needed, interpret the generated prediction results to ensure that medical professionals can understand and make decisions based on them, use prediction results and user feedback for continuous optimization and adjustment of the model, generate detailed test reports based on prediction results for reference by medical professionals and users, continuously monitor and track user health status, and ensure that any health problems are discovered and addressed in a timely manner. Medical professionals can obtain accurate and timely information about the user's health status, thereby providing users with more accurate and personalized medical services. At the same time, this process also supports clinical decision-making and helps improve the quality and efficiency of medical services.

[0068] The present application can combine dual continuous variables. For example, when establishing a reference interval for TSH, the factors of the dual continuous variables of age and month can be integrated to establish a continuous reference interval, such as Figure 5 As shown, Figure 5 The upper limit surface of the reference interval of a certain item is shown as a surface that changes with age and month. The blue dot surface is the upper limit fitting surface of the continuous reference interval of the target item with different ages and months (95 percentile fitting surface). In addition, this application integrates the influence of non-continuous variables such as gender, region, and continuous variables such as age, season (month), and BMI to establish a continuous reference interval, considering more factors, and the established reference interval is more specific and sensitive.

[0069] In this application, the server obtains the target user data; processes the target user data to generate the user data to be monitored with identification information, wherein the identification information is used to characterize the target item to be detected by the target user at present; obtains the total training set matching the identification information and the preset multivariate continuous biological reference interval model matching the identification information, wherein the total training set includes the training samples corresponding to the target item to be detected by the target user; pre-processes the total training set to generate the target training set and the target test set; trains the preset multivariate continuous biological reference interval model based on the target training set and the target test set to generate the target multivariate continuous biological reference interval model; processes the user data to be monitored based on the target multivariate continuous biological reference interval model to generate the prediction result, wherein the prediction result includes the detection status of the disease to be detected by the target user. By simultaneously considering non-continuous variables and continuous variables, a model of continuous reference intervals that can reflect multiple continuous variables is established, aiming to capture the dynamic changes of test indicators under normal physiological conditions, improve the accuracy of clinical tests, reduce misdiagnosis and missed diagnosis, provide support for personalized medicine, and enable clinicians to manage and treat diseases more effectively.

[0070] Optionally, in another embodiment of the method of the present application, the step of processing the user data to be monitored based on the target multivariate continuous biological reference interval model to generate a prediction result includes:

[0071] If the patient characteristic information indicates that the target user is a patient of the target hospital, obtaining the case information of the target user;

[0072] Acquire historical laboratory indicator data of the target user based on the case information of the target user;

[0073] The historical laboratory indicator data is processed based on the target multivariate continuous biological reference interval model to generate historical prediction results.

[0074] In one implementation, first, it is identified whether the target user is a patient of the target hospital based on the patient characteristic information. If the target user is a patient of the target hospital, the case information is obtained, including past medical history, diagnosis results, etc., and the historical laboratory indicator data of the target user is extracted from the case information, wherein the historical laboratory indicator data includes but is not limited to non-continuous variables such as gender, region, and continuous variables such as age, month, and BMI. The historical laboratory indicator data is integrated with the current biological reference interval model to facilitate historical data analysis.

[0075] The target multivariate continuous biological reference interval model is used to process historical laboratory indicator data. Based on the model processing results, historical prediction results for the target users are generated to ensure that the historical prediction results are consistent with the current biological reference interval model. Trend analysis is performed on the historical prediction results to identify changing trends in the user's health status. The target user's health risks and possible disease development are assessed based on the historical prediction results. A more comprehensive user health profile is provided by combining real-time vital sign information and historical prediction results.

[0076] Optionally, in another embodiment based on the above method of the present application, the total training set is preprocessed to generate a target training set and a target test set, further comprising:

[0077] Processing the target training set to generate an initial candidate group and an initial fixed group;

[0078] The initial candidate group and the initial fixed group are processed respectively to generate a preset candidate group and a preset fixed group, wherein the preset candidate group and the preset fixed group respectively carry identification information representing different covariate priority information;

[0079] Generate a loop iteration rule based on the identification information, wherein the loop iteration rule is set based on the number of covariates to be matched that is preset by a user;

[0080] Processing the covariates in the preset candidate group and the covariates in the preset fixed group based on the loop iteration rule to generate a preprocessed candidate group and a preprocessed fixed group, wherein the covariates include continuous variables and categorical variables;

[0081] Processing the preprocessing selected group and the preprocessing fixed group based on a preset matching rule to generate a matching result;

[0082] Processing the matching results based on a variable numerical calculation model to generate statistical values ​​of the preprocessed selected group and statistical values ​​of the preprocessed fixed group;

[0083] If the statistical value of the preprocessed candidate group and the statistical value of the preprocessed fixed group meet the matching requirements respectively, a target candidate group and a target fixed group are generated.

[0084] In one embodiment, the total training set is subjected to preprocessing operations such as cleaning and standardization to prepare for generating a target training set and a target test set, and the preprocessed data set is divided into two parts: a target training set for model training and a target test set for model evaluation. In the target training set, an initial candidate group and a fixed group are generated according to specific standards or rules. The initial candidate group and the fixed group are processed to generate a preset candidate group and a preset fixed group, and these groups carry identifiers of different covariate priority information. Based on the number of covariates set by the user, a loop iteration rule is generated for subsequent covariate processing. According to the loop iteration rule, the covariates (including continuous variables and categorical variables) in the preset candidate group and the preset fixed group are processed to generate a preprocessed candidate group and a preprocessed fixed group.

[0085] Based on the preset matching rules, the preprocessed candidate groups and fixed groups are matched to generate matching results. The matching results are processed using the variable numerical calculation model to generate statistical values ​​of the preprocessed candidate groups and preprocessed fixed groups. The statistical values ​​of the preprocessed candidate groups and preprocessed fixed groups are checked to see if they meet the matching requirements. If the statistical values ​​meet the requirements, these groups are defined as the target candidate groups and target fixed groups, which will be used for the final model training. The data of the target candidate groups and target fixed groups are used to prepare for model training and optimization. The performance of the model is evaluated on the target test set, and adjustments and optimizations are made as needed. The verified and optimized model is applied to actual problems for prediction and analysis. Through this process, the quality and representativeness of the model training data can be ensured, thereby improving the accuracy and reliability of the model. At the same time, this process also supports the interpretability and stability evaluation of the model.

[0086] By applying the above technical solution, the server obtains target user data; processes the target user data to generate user attribute information, wherein the user attribute information includes the target items to be detected by the target user; performs feature extraction on the target user data to generate a feature data set; processes the feature data set based on the original feature library to generate patient feature information; processes the patient feature information based on the user attribute information to generate user data to be monitored with identification information, wherein the identification information is used to characterize the target items to be detected by the target user at present.

[0087] Obtain a total training set that matches the identification information and a preset multivariate continuous biological reference interval model that matches the identification information, wherein the total training set includes training samples corresponding to the target items to be detected by the target user; perform feature extraction on the total training set to determine an original feature library, wherein the original feature library includes non-continuous variable data and continuous variable data; process the total training set based on the non-continuous variable data to generate an initial training set; process the initial training set based on the continuous variable data to generate a target training set, wherein the target training set includes several biological reference intervals, and different biological reference intervals correspond to different disease types; process the target training set to generate an initial candidate group and an initial fixed group; process the initial candidate group and the initial fixed group respectively to generate a preset candidate group and a preset fixed group, wherein the preset The candidate group and the preset fixed group respectively carry identification information representing different covariate priority information; a loop iteration rule is generated based on the identification information, wherein the loop iteration rule is set based on the number of covariates to be matched preset by the user; the covariates in the preset candidate group and the covariates in the preset fixed group are processed based on the loop iteration rule to generate a preprocessed candidate group and a preprocessed fixed group, wherein the covariates include continuous variables and categorical variables; the preprocessed candidate group and the preprocessed fixed group are processed based on the preset matching rule to generate a matching result; the matching result is processed based on a variable numerical calculation model to generate a statistical value of the preprocessed candidate group and a statistical value of the preprocessed fixed group; if the statistical value of the preprocessed candidate group and the statistical value of the preprocessed fixed group meet the matching requirements respectively, a target candidate group and a target fixed group are generated.

[0088] The target training set is processed to generate multiple continuous variable data matching the identification information; the multiple continuous variable data is processed based on a preset function to generate a target biological reference interval, wherein the target biological reference interval is a preset reference range matching the target item; a target multivariate continuous biological reference interval model is generated based on the target biological reference interval; if the patient characteristic information represents that the target user is a patient of the target hospital, the case information of the target user is obtained; the historical laboratory indicator data of the target user is obtained based on the case information of the target user; the historical laboratory indicator data is processed based on the target multivariate continuous biological reference interval model to generate historical prediction results, and the real-time laboratory indicator data of the user to be monitored is obtained; the laboratory indicator data is processed based on the target multivariate continuous biological reference interval model to generate real-time vital sign information of the user to be monitored; the real-time vital sign information of the user to be monitored is processed based on the historical prediction results to generate the detection results of the target item of the user to be monitored within a preset time period, wherein the prediction results include the detection status of the target disease to be detected by the target user. By considering both non-continuous and continuous variables, a model that can reflect the continuous reference intervals of multiple continuous variables is established, aiming to capture the dynamic changes of test indicators under normal physiological conditions, improve the accuracy of clinical tests, reduce misdiagnosis and missed diagnosis, provide support for personalized medicine, and enable clinicians to manage and treat diseases more effectively.

[0089] In one embodiment, if Figure 2 As shown, the present application also provides a device for generating a multi-dimensional dynamic continuous biological reference interval, comprising:

[0090] The acquisition module 201 is used to acquire target user data; acquire a total training set matching the identification information and a preset multivariate continuous biological reference interval model matching the identification information, wherein the total training set includes training samples corresponding to the target items to be detected for the target user;

[0091] Processing module 202 is used to process the target user data to generate user data to be monitored with identification information, wherein the identification information is used to characterize the target item to be detected by the target user at present; preprocess the total training set to generate a target training set and a target test set; train the preset multivariate continuous biological reference interval model based on the target training set and the target test set to generate a target multivariate continuous biological reference interval model; process the user data to be monitored based on the target multivariate continuous biological reference interval model to generate a prediction result, wherein the prediction result includes the detection status of the disease to be detected by the target user.

[0092] In another embodiment of the present application, the processing module 202 is configured to pre-process the total training set to generate a target training set and a target test set, including:

[0093] Performing feature extraction on the total training set to determine an original feature library, wherein the original feature library includes non-continuous variable data and continuous variable data;

[0094] Processing the total training set based on the non-continuous variable data to generate an initial training set;

[0095] The initial training set is processed based on the continuous variable data to generate a target training set, wherein the target training set includes a plurality of biological reference intervals, and different biological reference intervals correspond to different disease types.

[0096] In another embodiment of the present application, the processing module 202 is configured to train the preset multivariate continuous biological reference interval model based on the target training set and the target test set to generate a target multivariate continuous biological reference interval model, including:

[0097] Processing the target training set to generate multiple continuous variable data matching the identification information;

[0098] Processing the multiple continuous variable data based on a preset function to generate a target biological reference interval, wherein the target biological reference interval is a preset reference range that matches the target item;

[0099] A target multivariate continuous biological reference interval model is generated based on the target biological reference interval.

[0100] In another embodiment of the present application, the processing module 202 is configured to process the target user data to generate the user data to be monitored with identification information, including:

[0101] Processing the target user data to generate user attribute information, wherein the user attribute information includes a target item to be detected by the target user;

[0102] Extracting features from the target user data to generate a feature data set;

[0103] Processing the feature data set based on the original feature library to generate patient feature information;

[0104] The patient characteristic information is processed based on the user attribute information to generate user data to be monitored with identification information.

[0105] In another embodiment of the present application, the processing module 202 is configured to process the user data to be monitored based on the target multivariate continuous biological reference interval model to generate a prediction result, including:

[0106] If the patient characteristic information indicates that the target user is a patient of the target hospital, obtaining the case information of the target user;

[0107] Acquire historical laboratory indicator data of the target user based on the case information of the target user;

[0108] The historical laboratory indicator data is processed based on the target multivariate continuous biological reference interval model to generate historical prediction results.

[0109] In another embodiment of the present application, the processing module 202 is configured to process the user data to be monitored based on the target multivariate continuous biological reference interval model to generate a prediction result, and further includes:

[0110] Obtain real-time laboratory indicator data of the user to be monitored;

[0111] Processing the laboratory index data based on the target multivariate continuous biological reference interval model to generate real-time vital sign information of the user to be monitored;

[0112] The real-time vital sign information of the user to be monitored is processed based on the historical prediction results to generate detection results of target items of the user to be monitored within a preset time period.

[0113] In another embodiment of the present application, the processing module 202 is configured to pre-process the total training set to generate a target training set and a target test set, and further includes:

[0114] Processing the target training set to generate an initial candidate group and an initial fixed group;

[0115] The initial candidate group and the initial fixed group are processed respectively to generate a preset candidate group and a preset fixed group, wherein the preset candidate group and the preset fixed group respectively carry identification information representing different covariate priority information;

[0116] Generate a loop iteration rule based on the identification information, wherein the loop iteration rule is set based on the number of covariates to be matched that is preset by a user;

[0117] Processing the covariates in the preset candidate group and the covariates in the preset fixed group based on the loop iteration rule to generate a preprocessed candidate group and a preprocessed fixed group, wherein the covariates include continuous variables and categorical variables;

[0118] Processing the preprocessing selected group and the preprocessing fixed group based on a preset matching rule to generate a matching result;

[0119] Processing the matching results based on a variable numerical calculation model to generate statistical values ​​of the preprocessed selected group and statistical values ​​of the preprocessed fixed group;

[0120] If the statistical value of the preprocessed candidate group and the statistical value of the preprocessed fixed group meet the matching requirements respectively, a target candidate group and a target fixed group are generated.

[0121] The present application embodiment provides an electronic device, such as Figure 3 As shown, the electronic device 3 includes a first processor 300, a memory 301, a bus 302 and a communication interface 303. The first processor 300, the communication interface 303 and the memory 301 are connected via the bus 302; the memory 301 stores a computer program that can be run on the first processor 300, and when the first processor 300 runs the computer program, it executes the method for generating the multi-dimensional dynamic continuous biological reference interval provided in any of the aforementioned embodiments of the present application.

[0122] The memory 301 may include a high-speed random access memory (RAM), and may also include a non-volatile memory, such as at least one disk memory. The communication connection between the system network element and at least one other network element is realized through at least one communication interface 303 (which may be wired or wireless), and the Internet, wide area network, local area network, metropolitan area network, etc. may be used.

[0123] The bus 302 may be an ISA bus, a PCI bus, or an EISA bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. Among them, the memory 301 is used to store programs, and the first processor 300 executes the program after receiving the execution instruction. The method for generating the multi-dimensional dynamic continuous biological reference interval disclosed in any implementation of the embodiment of the present application may be applied to the first processor 300, or implemented by the first processor 300.

[0124] The first processor 300 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the hardware integrated logic circuit or software instructions in the first processor 300. The above-mentioned first processor 300 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a readily available programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in the embodiments of the present application can be embodied as a hardware decoding processor to execute, or a combination of hardware and software modules in the decoding processor to execute. The software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory 301, and the first processor 300 reads the information in the memory 301 and completes the steps of the above method in combination with its hardware.

[0125] The electronic device provided in the above-mentioned embodiments of the present application and the method for generating a multi-dimensional dynamic continuous biological reference interval provided in the embodiments of the present application are based on the same inventive concept and have the same beneficial effects as the method adopted, run or implemented by the application program stored therein.

[0126] The present application embodiment provides a computer-readable storage medium, such as Figure 4 As shown, the computer-readable storage medium stores 401 a computer program, and when the computer program is read and executed by the second processor 402, the method for generating a multi-dimensional dynamic continuous biological reference interval as described above is implemented.

[0127] The technical solution of the embodiment of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for enabling an electronic device (which may be an air conditioner, a refrigeration device, a personal computer, a server, or a network device, etc.) or a processor to execute all or part of the steps of the method described in the embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a ROM, a RAM, a magnetic disk, or an optical disk.

[0128] The computer-readable storage medium provided in the above-mentioned embodiments of the present application and the method for generating a multi-dimensional dynamic continuous biological reference interval provided in the embodiments of the present application are based on the same inventive concept and have the same beneficial effects as the method adopted, run or implemented by the application program stored therein.

[0129] An embodiment of the present application provides a computer program product, including a computer program, wherein the computer program is executed by a third processor to implement the method described above.

[0130] The computer program product provided in the above-mentioned embodiments of the present application and the method for generating a multi-dimensional dynamic continuous biological reference interval provided in the embodiments of the present application are based on the same inventive concept and have the same beneficial effects as the method adopted, run or implemented by the application program stored therein.

[0131] It should be noted that, in this application, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or still includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise one..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0132] Each embodiment in the present application is described in a related manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the generation method, electronic device, electronic device, and readable storage medium embodiment for evaluating multi-dimensional dynamic continuous biological reference intervals, since they are basically similar to the above-mentioned generation method embodiment of multi-dimensional dynamic continuous biological reference intervals, the description is relatively simple, and the relevant parts can be referred to the partial description of the above-mentioned generation method embodiment of multi-dimensional dynamic continuous biological reference intervals.

[0133] Although the present application is disclosed as above, the present application is not limited thereto. Any person skilled in the art can make various changes and modifications without departing from the spirit and scope of the present application, so the protection scope of the present application shall be subject to the scope defined by the claims.

Claims

1. A method for generating a multi-dimensional dynamic continuous biological reference interval, characterized in that: include: Obtain target user data; Processing the target user data to generate user data to be monitored with identification information, wherein the identification information is used to characterize the target item to be detected by the target user; Acquire a total training set matching the identification information and a preset multivariate continuous biological reference interval model matching the identification information, wherein the total training set includes training samples corresponding to target items to be detected for the target user; Preprocessing the total training set to generate a target training set and a target test set; Training the preset multivariate continuous biological reference interval model based on the target training set and the target test set to generate a target multivariate continuous biological reference interval model; The user data to be monitored is processed based on the target multivariate continuous biological reference interval model to generate a prediction result, wherein the prediction result includes the detection status of the disease to be detected of the target user.

2. The method according to claim 1, characterized in that The preprocessing of the total training set to generate a target training set and a target test set includes: Performing feature extraction on the total training set to determine an original feature library, wherein the original feature library includes non-continuous variable data and continuous variable data; Processing the total training set based on the non-continuous variable data to generate an initial training set; The initial training set is processed based on the continuous variable data to generate a target training set, wherein the target training set includes a plurality of biological reference intervals, and different biological reference intervals correspond to different disease types.

3. The method according to claim 1, characterized in that The step of training the preset multivariate continuous biological reference interval model based on the target training set and the target test set to generate a target multivariate continuous biological reference interval model includes: Processing the target training set to generate multiple continuous variable data matching the identification information; Processing the multiple continuous variable data based on a preset function to generate a target biological reference interval, wherein the target biological reference interval is a preset reference range that matches the target item; A target multivariate continuous biological reference interval model is generated based on the target biological reference interval.

4. The method according to claim 1, characterized in that The processing of the target user data to generate the user data to be monitored with identification information includes: Processing the target user data to generate user attribute information, wherein the user attribute information includes a target item to be detected by the target user; Extracting features from the target user data to generate a feature data set; Processing the feature data set based on the original feature library to generate patient feature information; The patient characteristic information is processed based on the user attribute information to generate user data to be monitored with identification information.

5. The method according to claim 4, characterized in that The processing of the user data to be monitored based on the target multivariate continuous biological reference interval model to generate a prediction result includes: If the patient characteristic information indicates that the target user is a patient of the target hospital, obtaining the case information of the target user; Acquire historical laboratory indicator data of the target user based on the case information of the target user; The historical laboratory indicator data is processed based on the target multivariate continuous biological reference interval model to generate historical prediction results.

6. The method according to claim 5, characterized in that The processing of the user data to be monitored based on the target multivariate continuous biological reference interval model to generate a prediction result also includes: Obtain real-time laboratory indicator data of the user to be monitored; Processing the laboratory index data based on the target multivariate continuous biological reference interval model to generate real-time vital sign information of the user to be monitored; The real-time vital sign information of the user to be monitored is processed based on the historical prediction results to generate detection results of target items of the user to be monitored within a preset time period.

7. The method according to claim 1, characterized in that Preprocessing the total training set to generate a target training set and a target test set also includes: Processing the target training set to generate an initial candidate group and an initial fixed group; The initial candidate group and the initial fixed group are processed respectively to generate a preset candidate group and a preset fixed group, wherein the preset candidate group and the preset fixed group respectively carry identification information representing different covariate priority information; Generate a loop iteration rule based on the identification information, wherein the loop iteration rule is set based on the number of covariates to be matched that is preset by a user; Processing the covariates in the preset candidate group and the covariates in the preset fixed group based on the loop iteration rule to generate a preprocessed candidate group and a preprocessed fixed group, wherein the covariates include continuous variables and categorical variables; Processing the preprocessing selected group and the preprocessing fixed group based on a preset matching rule to generate a matching result; Processing the matching results based on a variable numerical calculation model to generate statistical values ​​of the preprocessed selected group and statistical values ​​of the preprocessed fixed group; If the statistical value of the preprocessed candidate group and the statistical value of the preprocessed fixed group meet the matching requirements respectively, a target candidate group and a target fixed group are generated.

8. A device for generating a multi-dimensional dynamic continuous biological reference interval, characterized in that: The device comprises: An acquisition module is used to acquire target user data; acquire a total training set matching the identification information and a preset multivariate continuous biological reference interval model matching the identification information, wherein the total training set includes training samples corresponding to target items to be detected for the target user; A processing module is used to process the target user data to generate user data to be monitored with identification information, wherein the identification information is used to characterize the target item to be detected by the target user at present; preprocess the total training set to generate a target training set and a target test set; train the preset multivariate continuous biological reference interval model based on the target training set and the target test set to generate a target multivariate continuous biological reference interval model; process the user data to be monitored based on the target multivariate continuous biological reference interval model to generate a prediction result, wherein the prediction result includes the detection status of the disease to be detected by the target user.

9. An electronic device, characterized in that: include: a first processor; and a memory for storing executable instructions of the first processor; Wherein, the first processor is configured to execute the method for generating a multi-dimensional dynamic continuous biological reference interval as described in any one of claims 1 to 7 by executing the executable instructions.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the second processor, the method for generating a multi-dimensional dynamic continuous biological reference interval as described in any one of claims 1 to 7 is implemented.

Images