Method for predicting disease condition group of patient

Through the preprocessing of patient data and mathematical statistical algorithm models, the problem of failure to predict treatment response and disease progress in the prior art is solved, and accurate prediction of patient condition groups is achieved, and treatment effect and efficiency are improved.

CN118919061APending Publication Date: 2024-11-08THE SEVENTH AFFILIATED HOSPITAL SUN YAT SEN UNIV SHENZHEN

Patent Information

Application Number
CN202411093613.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-09
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

Existing prediction methods do not consider grouping patients to predict their treatment response and disease progression, resulting in the inability to effectively evaluate and develop personalized treatment options.

Method used

By collecting and preprocessing the patient's basic information and timing detection data, the rate of decline of HBsAg levels relative to baseline is calculated, and a prediction model is established using mathematical statistical algorithms to automatically analyze and predict the potential category groups to which the patient belongs.

Benefits of technology

It improves data quality and prediction accuracy, can quickly and accurately predict patients' condition groups, helping doctors formulate personalized treatment plans, improve treatment effects and efficiency, and save time and resources.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118919061A_ABST
    Figure CN118919061A_ABST
Patent Text Reader

Abstract

The invention relates to a method for predicting disease condition groups of patients, and belongs to the field of medical data analysis. The invention discloses a method for predicting disease condition groups of patients. The method comprises the following steps: S1, collecting basic information and time sequence detection data of the patients; s2, preprocessing the data of the patient, wherein the preprocessing comprises data cleaning and missing value processing; s3, calculating the decrease rate of the HBsAg level of the patient at each detection moment relative to the baseline; s4, predicting a potential category group to which the patient belongs based on a calculation result; and S5, outputting a prediction result. By preprocessing the patient data, the accuracy of the data is ensured, the decrease rate of the HBsAg level of the patient at each detection moment relative to the baseline is calculated, and the change of the disease condition of the patient is accurately known. And on the basis of the calculation result of the HBsAg level decrease rate, the potential category group to which the patient belongs is predicted, so that a doctor can formulate a treatment scheme and the patient can know the health condition of himself / herself. By automatically processing and analyzing a large amount of data and quickly outputting a prediction result, time and energy of doctors and patients are saved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of medical data analysis, and in particular to a method for predicting patient condition groups. Background Art

[0002] Chronic hepatitis B is an important public health issue, and the prediction of its treatment response and disease progression is critical for the treatment and care of patients. In particular, changes in the level of HBsAg (hepatitis B surface antigen) have become an important indicator for evaluating the course of the disease and the effectiveness of treatment. However, existing prediction methods do not consider grouping patients to predict their treatment response and disease progression. Therefore, a method for predicting the patient's condition group is urgently needed. Summary of the invention

[0003] Based on this, it is necessary to provide a method for predicting patient condition groups to address the problem that existing prediction methods do not group patients to predict their treatment response and disease progression.

[0004] A method for predicting a patient's condition group, comprising the following steps:

[0005] S1. Collect the patient's basic information and time series test data;

[0006] S2. Preprocess the patient's data, including data cleaning and missing value processing, to obtain analytical data;

[0007] S3, calculating the decline rate of the patient's HBsAg level relative to the baseline at each testing time;

[0008] S4, predicting the latent class group to which the patient belongs based on the calculation results;

[0009] S5. Output prediction results.

[0010] The present application discloses a method for predicting the patient's condition group. By preprocessing the patient data, including data cleaning and missing value processing, the data quality can be greatly improved, the accuracy and completeness of the data can be ensured, and a reliable data basis can be provided for subsequent analysis and prediction. By calculating the decline rate of the patient's HBsAg level relative to the baseline at each detection moment, the patient's condition changes can be understood more accurately, and the precise calculation method helps to more accurately assess the patient's condition. Based on the data collected and processed in the above steps and the calculation results of the HBsAg level decline rate, the potential category group to which the patient belongs can be accurately predicted, which is of great significance for doctors to formulate treatment plans and patients to understand their health conditions. In addition, the method can quickly output the prediction results by automatically processing and analyzing a large amount of data, thereby predicting the potential category group to which the patient's condition belongs, greatly improving the efficiency of the prediction and saving the time and energy of doctors and patients. In summary, the method for predicting the patient's condition group effectively realizes the prediction of the patient's condition group by improving data quality, accurate calculation, improving prediction accuracy, and being efficient and convenient, and has important practical value and significance for the medical field.

[0011] In any of the above technical solutions, step S4 includes the following steps:

[0012] S41, selecting a first feature related to treatment response and disease progression from the pre-processed data, wherein the first feature includes a rate of decrease in HBsAg level;

[0013] S42, using mathematical statistics algorithms to establish a prediction model and divide the analysis data into a training data set and a test data set;

[0014] S43, training the model using the training data set to optimize model parameters;

[0015] S44, using the test data set to evaluate the model and evaluate the prediction performance of the model;

[0016] S45, inputting the patient's information into the prediction model, wherein the patient's information includes the HBsAg level decrease rate;

[0017] S46. According to the prediction results output by the model, the group with the largest posterior probability in the prediction results is selected as the latent category group to which the patient belongs.

[0018] By using mathematical statistics algorithms to establish a prediction model, the accuracy and reliability of the prediction can be improved, which helps to improve the treatment effect and survival rate of patients. Using training data sets to train and optimize the model, and using test data sets to evaluate the model, the prediction performance of the model can be verified and the reliability of the model can be ensured. Inputting the calculated HBsAg level decline rate and other key information into the prediction model can quickly and accurately predict the patient's disease group, which helps to take appropriate treatment measures in a timely manner. According to the prediction results output by the model, the group with the largest posterior probability is selected as the potential category group to which the patient belongs, which can improve the accuracy and credibility of the prediction and help guide doctors to carry out personalized treatment.

[0019] In any of the above technical solutions, let Y i ={y i,t} represents the rate of decrease of HBsAg level of patient i at the tth detection time relative to the baseline. The posterior probability of the patient belonging to the jth group is calculated by the following formula:

[0020]

[0021] Where T is the maximum number of weeks that the patient has been tested for that needs to be predicted; j = 1, 2, 3;

[0022] The proportion of patients in each group π={π j} is given by:

[0023]

[0024] Among them, each group of parameters θ=(θ1,θ2,θ3)=(0.2021791,0,-0.373173), e is a natural constant, also known as Euler number, and its value is 2.71828.

[0025] Among them, the actual response variable is censored, so p j (y i,t ) is given by:

[0026]

[0027] Among them, y min =-2,y max =1, variance σ = (σ1, σ2, σ3) = (0.17052475, 0.28058448, 0.30667541), Φ(x) is the standard normal cumulative distribution function, and its calculation formula is y i,t represents the decline rate of HBsAg level relative to the baseline of patient i at the tth detection time, is the predicted rate of decrease in HBsAg level relative to the baseline when patient i belongs to trajectory group j at the tth detection time, which is given by:

[0028]

[0029] Among them, w i,t It indicates the number of weeks after the baseline test for this patient at the tth test time.

[0030] By using the prediction model to calculate the posterior probability of the group to which the patient belongs, the patient's condition group can be accurately predicted. This method takes into account the decline rate of the patient's HBsAg level relative to the baseline at each detection time, as well as other related factors, thereby improving the accuracy of the prediction. The prediction model is used to calculate the posterior probability, which fully considers the individual differences of patients, including the speed and pattern of their disease onset, which makes the prediction results more personalized and more reflective of the actual situation of the patients. By introducing the predicted decline rate of the HBsAg level of patients in the trajectory group relative to the baseline at the tth detection time, the limitations of the actual dependent variable can be handled, which helps to more accurately evaluate the changes in the patient's condition and make more accurate group predictions based on this. The use of specific calculation formulas and parameters simplifies the prediction process, making the prediction operation more convenient and efficient, helping doctors to make a quick diagnosis and provide patients with timely treatment plans. By using mathematical statistics algorithms to establish a prediction model, intelligent medical decision support can be achieved, which helps doctors analyze patient data more quickly and improve the accuracy of diagnosis.

[0031] In any of the above technical solutions, in step S42, a prediction model is established using a mathematical statistics algorithm, including the following steps:

[0032] S421. Optimize the parameters of mathematical statistics algorithms to improve the prediction performance of the model;

[0033] S422, perform cross validation to evaluate the generalization ability of the model;

[0034] S423. Perform model selection to select the prediction model that best suits the patient data.

[0035] By tuning the parameters of mathematical statistics algorithms, data features can be matched more accurately, thereby improving the performance of the prediction model, helping to more accurately predict the patient's condition or behavior and providing strong support for medical decision-making. Through cross-validation, the performance of the prediction model on unknown data can be evaluated to determine the model's ability to be optimized, which helps ensure the stability and reliability of the model in practical applications and reduce the risk of overfitting and underfitting. Model selection can find the prediction model that best suits a specific patient's data from multiple models, helping to ensure that the selected model can maximize the characteristics of the patient's data and improve the accuracy of the prediction.

[0036] In any of the above technical solutions, in step S5, the prediction result includes:

[0037] The decline rate trend chart of HBsAg level at different prediction weeks relative to the baseline;

[0038] The latent class group to which the patient belongs.

[0039] By graphically displaying the trend of the decline rate of HBsAg levels relative to the baseline in different prediction weeks, doctors can more intuitively and accurately understand the trend of patients' condition changes, thereby providing patients with more personalized treatment plans. By predicting the potential category group to which the patient belongs (rapid high clearance group, delayed high clearance group and slow low clearance group), doctors can more quickly identify patients with better treatment effects, thereby optimizing resource allocation and improving overall treatment efficiency, or find patients with poor treatment effects earlier, adjust treatment plans in time, avoid unnecessary treatment delays and waste of resources, and thus reduce medical costs. In addition, medical institutions can accumulate valuable medical data by collecting and analyzing patients' prediction results, provide reference for future research and treatment, and promote the development of the medical field.

[0040] In any of the above technical solutions, based on the decline rate data of the patient's HBsAg level relative to the baseline at each detection time, the change trend of the HBsAg level relative to the baseline in different prediction weeks is graphically displayed, and the potential category group to which the patient belongs in different prediction weeks is quantitatively displayed through the posterior probability.

[0041] By displaying the decline rate data of the patient's HBsAg level relative to the baseline at each test moment in a graphical manner, the change trend of the HBsAg level relative to the baseline in different forecast weeks can help doctors and patients understand the patient's disease progression more intuitively and predict the future change trend of HBsAg levels. In this way, doctors can more accurately assess the patient's disease progression and develop personalized treatment plans. At the same time, patients can also better understand their own condition, enhance their confidence in treatment, and help improve treatment effects.

[0042] In any of the above technical solutions, in step S1, the basic information includes gender, age and BMI, the time series detection data includes detection date, HBsAg level, ALT level and HBV DNA level, and the ALT level and the HBV DNA level are optional items.

[0043] By introducing basic information such as gender, age and BMI, as well as time-series test data such as test date and HBsAg level, the patient's health status can be fully and deeply understood and analyzed, and the patient's liver function and hepatitis B virus infection can be monitored and tracked in real time. The optional ALT (alanine aminotransferase) level and HBV DNA (hepatitis B virus) level provide individuals with more flexible choices, and they can determine whether to conduct corresponding tests based on their specific circumstances. By collecting and analyzing this information, patients can better understand their health status, detect potential health problems in a timely manner, and take corresponding preventive and treatment measures.

[0044] In any of the above technical solutions, the basic information and the time series detection data are collected through a medical information system and preprocessed, and the preprocessing includes data cleaning and missing value processing.

[0045] Data collection and preprocessing through medical information systems can ensure the quality of the collected basic information and time series detection data. Medical information systems can automatically process large amounts of data, reduce the tediousness of manual operations, and improve decision-making efficiency, which is very important for the daily operations of medical institutions and the treatment process of patients. Data cleaning and missing value processing improve the accuracy and reliability of data. Preprocessed data can more accurately reflect the actual situation of patients, help doctors make more accurate diagnoses and treatment plans, and thus improve treatment effects.

[0046] In any of the above technical solutions, in step S3, the descent rate is calculated by the following formula:

[0047]

[0048] Among them, x 4,i,0 represents the baseline HBsAg level, x 4,i,t represents the HBsAg level at the tth detection time, y i,t It represents the decrease rate of HBsAg level of patient i at the tth detection time relative to the baseline.

[0049] By introducing the baseline HBsAg level and the HBsAg level at the tth detection time, the decline rate of the patient's HBsAg level relative to the baseline at the tth detection time is calculated, making the prediction result closer to the actual situation. Since a linear model is used, a reasonable truncation mechanism is set after the prediction to avoid the prediction result not being in line with the actual situation and improve the stability and reliability of the model.

[0050] In any of the above technical solutions, in step S5, the prediction result further includes:

[0051] Follow up the patient's condition over a long period of time to monitor changes in the patient's condition;

[0052] Communicate the patient's medical condition information and predicted results with medical experts to obtain more professional treatment advice.

[0053] By tracking the patient's condition for a long time, changes in the condition can be discovered in time, providing an important basis for subsequent treatment; the patient's condition information and predicted results can be communicated with medical experts to obtain more professional treatment suggestions, thereby improving the pertinence and effectiveness of treatment, avoiding unnecessary repeated examinations and treatments, reducing the waste of medical resources, and improving the utilization efficiency of medical resources. In addition, a large amount of patient condition information and treatment experience can be accumulated, providing valuable data support for the development of medical technology and promoting the advancement of medical technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 A flowchart of the steps of a method for predicting a patient's condition group;

[0055] Figure 2 Flowchart of the steps to predict the latent class group to which a patient belongs;

[0056] Figure 3 Flowchart of the steps for building a predictive model. DETAILED DESCRIPTION

[0057] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0058] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0059] A method for predicting a patient's condition group according to some embodiments of the present invention will be described below with reference to the accompanying drawings.

[0060] like Figure 1 As shown, this embodiment discloses a method for predicting the patient's condition group, and the specific steps are as follows:

[0061] S1. Collect the patient's basic information and time series test data;

[0062] S2. Preprocess the patient's data, including data cleaning and missing value processing, to obtain analytical data;

[0063] S3, calculating the decline rate of the patient's HBsAg level relative to the baseline at each testing time;

[0064] S4, predicting the latent class group to which the patient belongs based on the calculation results;

[0065] S5. Output prediction results.

[0066] The present application discloses a method for predicting the patient's condition group. By preprocessing the patient data, including data cleaning and missing value processing, the data quality can be greatly improved, the accuracy and completeness of the data can be ensured, and a reliable data basis can be provided for subsequent analysis and prediction. By calculating the decline rate of the patient's HBsAg level relative to the baseline at each detection moment, the patient's condition changes can be understood more accurately, and the precise calculation method helps to more accurately assess the patient's condition. Based on the data collected and processed in the above steps and the calculation results of the HBsAg level decline rate, the potential category group to which the patient belongs can be accurately predicted, which is of great significance for doctors to formulate treatment plans and patients to understand their health status. In addition, the method can quickly output the prediction results by automatically processing and analyzing a large amount of data, thereby predicting the potential category group to which the patient's condition belongs, greatly improving the efficiency of the prediction, and saving the time and energy of doctors and patients. In summary, the method for predicting the patient's condition group effectively realizes the prediction of the patient's condition group by improving data quality, accurate calculation, improving prediction accuracy, and being efficient and convenient, and has important practical value and significance for the medical field.

[0067] like Figure 2 As shown, in any of the above technical solutions, step S4 includes the following steps:

[0068] S41, selecting a first feature related to treatment response and disease progression from the pre-processed data, the first feature comprising a rate of decrease in HBsAg level;

[0069] S42, using mathematical statistics algorithms to establish a prediction model and divide the analysis data into a training data set and a test data set;

[0070] S43, training the model using the training data set to optimize model parameters;

[0071] S44, using the test data set to evaluate the model and evaluate the prediction performance of the model;

[0072] S45, inputting the patient's information into the prediction model, wherein the patient's information includes the HBsAg level decrease rate;

[0073] S46. According to the prediction results output by the model, the group with the largest posterior probability in the prediction results is selected as the latent category group to which the patient belongs.

[0074] By using mathematical statistics algorithms to establish a prediction model, the accuracy and reliability of the prediction can be improved, which helps to improve the treatment effect and survival rate of patients. Using training data sets to train and optimize the model, and using test data sets to evaluate the model, the prediction performance of the model can be verified and the reliability of the model can be ensured. Inputting the calculated BsAg level decline rate and other key information into the prediction model can quickly and accurately predict the patient's disease group, which helps to take appropriate treatment measures in a timely manner. According to the prediction results output by the model, the group with the largest posterior probability is selected as the potential category group to which the patient belongs, which can improve the accuracy and credibility of the prediction and help guide doctors to carry out personalized treatment.

[0075] In any of the above technical solutions, let Y i ={y i,t} represents the rate of decrease of HBsAg level of patient i at the tth detection time relative to the baseline. The posterior probability of the patient belonging to the jth group is calculated by the following formula:

[0076]

[0077] Where T is the maximum number of weeks that the patient to be predicted has been tested so far; j = 1, 2, 3;

[0078] The proportion of patients in each group π={π j} is given by:

[0079]

[0080] Among them, each group of parameters θ=(θ1,θ2,θ3)=(0.2021791,0,-0.373173), e is a natural constant, also known as Euler number, and its value is 2.71828.

[0081] Among them, the actual response variable is censored, so p j (y i,t ) is given by:

[0082]

[0083] Among them, y min =-2,y max =1, variance σ = (σ1, σ2, σ3) = (0.17052475, 0.28058448, 0.30667541), Φ(x) is the standard normal cumulative distribution function, and its calculation formula is y i,t represents the decline rate of HBsAg level relative to the baseline of patient i at the tth detection time, is the predicted rate of decrease in HBsAg level relative to the baseline when patient i belongs to trajectory group j at the tth detection time, which is given by:

[0084]

[0085] Among them, w i,t It indicates the number of weeks after the baseline test for this patient at the tth test time.

[0086] In the actual test, the predicted decrease rate data of HBsAg level relative to baseline of patient i belonging to different trajectory groups at different testing weeks are shown in Table 1.

[0087] By using the prediction model to calculate the posterior probability of the group to which the patient belongs, the patient's condition group can be accurately predicted. This method takes into account the decline rate of the patient's HBsAg level relative to the baseline at each detection time, as well as other related factors, thereby improving the accuracy of the prediction. The prediction model is used to calculate the posterior probability, which fully considers the individual differences of patients, including the speed and pattern of their disease onset, which makes the prediction results more personalized and more reflective of the actual situation of the patients. By introducing the predicted decline rate of the HBsAg level of patients in the trajectory group relative to the baseline at the tth detection time, the limitations of the actual dependent variable can be handled, which helps to more accurately evaluate the changes in the patient's condition and make more accurate group predictions based on this. The use of specific calculation formulas and parameters simplifies the prediction process, making the prediction operation more convenient and efficient, helping doctors to make a quick diagnosis and provide patients with timely treatment plans. By using mathematical statistics algorithms to establish a prediction model, intelligent medical decision support can be achieved, which helps doctors analyze patient data more quickly and improve the accuracy of diagnosis.

[0088] Table 1: Predicted decline rate of HBsAg level from baseline in each trajectory group

[0089]

[0090]

[0091]

[0092] Note: The labels of groups 1, 2, and 3 can be fast high-clearance group, delayed high-clearance group, and slow low-clearance group, respectively.

[0093] like Figure 3 As shown, in any of the above technical solutions, in step S42, a prediction model is established using a mathematical statistics algorithm, including the following steps:

[0094] S421. Optimize the parameters of mathematical statistics algorithms to improve the prediction performance of the model;

[0095] S422, perform cross validation to evaluate the generalization ability of the model;

[0096] S423. Perform model selection to select the prediction model that best suits the patient data.

[0097] By tuning the parameters of mathematical statistics algorithms, data features can be matched more accurately, thereby improving the performance of the prediction model, helping to more accurately predict the patient's condition or behavior and providing strong support for medical decision-making. Through cross-validation, the performance of the prediction model on unknown data can be evaluated to determine the model's ability to be optimized, which helps ensure the stability and reliability of the model in practical applications and reduce the risk of overfitting and underfitting. Model selection can find the prediction model that best suits a specific patient's data from multiple models, helping to ensure that the selected model can maximize the characteristics of the patient's data and improve the accuracy of the prediction.

[0098] In any of the above technical solutions, in step S5, the prediction result includes:

[0099] Trend graph of the decline rate of HBsAg level relative to the baseline at different prediction weeks;

[0100] The latent class group to which the patient belongs.

[0101] By graphically displaying the trend of the decline rate of HBsAg levels relative to the baseline in different prediction weeks, doctors can more intuitively and accurately understand the trend of patients' condition changes, thereby providing patients with more personalized treatment plans. By predicting the potential category group to which the patient belongs (rapid high clearance group, delayed high clearance group and slow low clearance group), doctors can more quickly identify patients with better treatment effects, thereby optimizing resource allocation and improving overall treatment efficiency, or find patients with poor treatment effects earlier, adjust treatment plans in time, avoid unnecessary treatment delays and waste of resources, and thus reduce medical costs. In addition, medical institutions can accumulate valuable medical data by collecting and analyzing patients' prediction results, provide reference for future research and treatment, and promote the development of the medical field.

[0102] In any of the above technical solutions, based on the decline rate data of the patient's HBsAg level relative to the baseline at each detection time, the change trend of the HBsAg level relative to the baseline in different prediction weeks is graphically displayed, and the potential category group to which the patient belongs in different prediction weeks is quantitatively displayed through the posterior probability.

[0103] By displaying the decline rate data of the patient's HBsAg level relative to the baseline at each test moment in a graphical manner, the change trend of the HBsAg level relative to the baseline in different forecast weeks can help doctors and patients understand the patient's disease progression more intuitively and predict the future change trend of HBsAg levels. In this way, doctors can more accurately assess the patient's disease progression and develop personalized treatment plans. At the same time, patients can also better understand their own condition, enhance their confidence in treatment, and help improve treatment effects.

[0104] In any of the above technical solutions, in step S1, the basic information includes gender, age and BMI, and the time series detection data includes detection date, HBsAg level, ALT level and HBV DNA level, and ALT level and HBV DNA level are optional items.

[0105] By introducing basic information such as gender, age and BMI, as well as time-series test data such as test date and HBsAg level, the patient's health status can be fully and deeply understood and analyzed, and the patient's liver function and hepatitis B virus infection can be monitored and tracked in real time. The optional ALT (alanine aminotransferase) level and HBV DNA (hepatitis B virus) level provide individuals with more flexible choices, and they can determine whether to conduct corresponding tests based on their specific circumstances. By collecting and analyzing this information, patients can better understand their health status, detect potential health problems in a timely manner, and take corresponding preventive and treatment measures.

[0106] In any of the above technical solutions, basic information and time series detection data are collected through a medical information system and preprocessed, and the preprocessing includes data cleaning and missing value processing.

[0107] Data collection and preprocessing through medical information systems can ensure the quality of the collected basic information and time series detection data. Medical information systems can automatically process large amounts of data, reduce the tediousness of manual operations, and improve decision-making efficiency, which is very important for the daily operations of medical institutions and the treatment process of patients. Data cleaning and missing value processing improve the accuracy and reliability of data. Preprocessed data can more accurately reflect the actual situation of patients, help doctors make more accurate diagnoses and treatment plans, and thus improve treatment effects.

[0108] In any of the above technical solutions, in step S3, the descent rate is calculated by the following formula:

[0109]

[0110] Among them, x 4,i,0 represents the baseline HBsAg level, x 4,i,trepresents the HBsAg level at the tth detection time week, y i,t It represents the decrease rate of HBsAg level of patient i at the tth detection time relative to the baseline.

[0111] By introducing the baseline HBsAg level and the HBsAg level at the tth detection time, the decline rate of the patient's HBsAg level relative to the baseline at the tth detection time is calculated, making the prediction result closer to the actual situation. Predicting the HBsAg level of the patient at a specific time point can timely predict the decline rate of the HBsAg level, helping doctors to identify changes in the condition early and take corresponding intervention measures. Since a linear model is used, setting a reasonable truncation mechanism after the prediction can avoid the prediction results not being in line with the actual situation and improve the stability and reliability of the model.

[0112] In any of the above technical solutions, in step S5, the prediction result further includes:

[0113] Follow up the patient's condition over a long period of time to monitor changes in the patient's condition;

[0114] Communicate the patient's medical condition information and predicted results with medical experts to obtain more professional treatment advice.

[0115] By tracking the patient's condition for a long time, changes in the condition can be discovered in time, providing an important basis for subsequent treatment; the patient's condition information and predicted results can be communicated with medical experts to obtain more professional treatment suggestions, thereby improving the pertinence and effectiveness of treatment, avoiding unnecessary repeated examinations and treatments, reducing the waste of medical resources, and improving the utilization efficiency of medical resources. In addition, a large amount of patient condition information and treatment experience can be accumulated, providing valuable data support for the development of medical technology and promoting the advancement of medical technology.

[0116] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0117] The above-mentioned embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A method for predicting a patient's condition group, characterized in that: The specific steps are as follows: S1. Collect the patient's basic information and time series test data; S2. Preprocess the patient's data, including data cleaning and missing value processing, to obtain analytical data; S3, calculating the decline rate of the patient's HBsAg level relative to the baseline at each testing time; S4, predicting the latent class group to which the patient belongs based on the calculation results; S5. Output prediction results.

2. The method for predicting a patient's condition group according to claim 1, characterized in that: The step S4 comprises the following steps: S41, selecting a first feature related to treatment response and disease progression from the pre-processed data, wherein the first feature includes a rate of decrease in HBsAg level; S42, using mathematical statistics algorithms to establish a prediction model and divide the analysis data into a training data set and a test data set; S43, training the model using the training data set to optimize model parameters; S44, using the test data set to evaluate the model and evaluate the prediction performance of the model; S45, inputting the patient's information into the prediction model, wherein the patient's information includes the HBsAg level decrease rate; S46. According to the prediction results output by the model, the group with the largest posterior probability in the prediction results is selected as the latent category group to which the patient belongs.

3. The method for predicting a patient's condition group according to claim 2, characterized in that: Let Y i ={y i,t } represents the rate of decrease of HBsAg level of patient i at the tth detection time relative to the baseline. The posterior probability of the patient belonging to the jth group is calculated by the following formula: Where T is the maximum number of weeks that the patient has been tested for that needs to be predicted; j = 1, 2, 3; The proportion of patients in each group π={π j } is given by: Among them, each group of parameters θ=(θ1,θ2,θ3)=(0.2021791,0,-0.373173), e is a natural constant, also known as Euler number, and its value is 2.71828; Among them, the actual response variable is censored, so p j (y i,t ) is given by: Among them, y min =-2,y max =1, variance σ = (σ1, σ2, σ3) = (0.17052475, 0.28058448, 0.30667541), Φ(x) is the standard normal cumulative distribution function, and its calculation formula is y i,t represents the decline rate of HBsAg level relative to the baseline of patient i at the tth detection time, is the predicted rate of decrease in HBsAg level relative to the baseline when patient i belongs to trajectory group j at the tth detection time, which is given by: Among them, w i,t It indicates the number of weeks after the baseline test for this patient at the tth test time.

4. The method for predicting a patient's condition group according to claim 2, characterized in that: In step S42, a prediction model is established using a mathematical statistics algorithm, including the following steps: S421. Optimize the parameters of mathematical statistics algorithms to improve the prediction performance of the model; S422, perform cross validation to evaluate the generalization ability of the model; S423. Perform model selection to select the prediction model that best suits the patient data.

5. The method for predicting a patient's condition group according to claim 1, characterized in that: In step S5, the prediction result includes: Trend graph of the decline rate of HBsAg level relative to the baseline at different prediction weeks; The latent class group to which the patient belongs.

6. The method for predicting a patient's condition group according to claim 5, characterized in that: Based on the data of the decline rate of the patient's HBsAg level relative to the baseline at each testing time, the changing trend of the HBsAg level relative to the baseline in different prediction weeks is displayed graphically, and the potential category group to which the patient belongs in different prediction weeks is quantitatively displayed through the posterior probability.

7. The method for predicting a patient's condition group according to claim 1, characterized in that: In step S1, the basic information includes gender, age and BMI, and the time series detection data includes detection date, HBsAg level, ALT level and HBV DNA level, and the ALT level and HBV DNA level are optional items.

8. The method for predicting a patient's condition group according to claim 1, characterized in that: The basic information and the time series detection data are collected through a medical information system and preprocessed, and the preprocessing includes data cleaning and missing value processing.

9. The method for predicting a patient's condition group according to claim 1, characterized in that: In step S3, the descent rate is calculated by the following formula: Among them, x 4,i,0 represents the baseline HBsAg level, x 4,i,t represents the HBsAg level at the tth detection time, y i,t It represents the decrease rate of HBsAg level of patient i at the tth detection time relative to the baseline.

10. The method for predicting a patient's condition group according to claim 1, characterized in that: In step S5, the prediction result further includes: Follow up the patient's condition over a long period of time to monitor changes in the patient's condition; Communicate the patient's medical condition information and predicted results with medical experts to obtain more professional treatment advice.

Citation Information

Patent Citations

  • Medical data processing method and device and storage medium

    CN112151136A

  • System and method for classifying progression of amyotrophic lateral sclerosis disease

    CN116052892A

  • Method and terminal for constructing prediction model of curative effect of transaminase normal chronic hepatitis B

    CN117457225A

  • Hepatitis B disease classification method

    CN117995382A

  • Method, system and device for evaluating human body index data in COVID-19 period and storage medium

    CN118248326A

Cited By

  • Method for predicting disease decline rate through mixed effect model

    CN118919060A

  • A method for predicting disease decline rate using mixed effects models

    CN118919060B