Search

CN-122000091-A - Method, device, medium and program product for predicting sirolimus dosing of children

CN122000091ACN 122000091 ACN122000091 ACN 122000091ACN-122000091-A

Abstract

The embodiment of the specification provides a method, equipment, medium and program product for predicting sirolimus administration dosage of children, and relates to the field of intelligent medical treatment. The method comprises the steps of obtaining predicted parameter information of a sample to be detected, wherein predicted parameters comprise weight, age, platelet count and creatinine, inputting the predicted parameter information into a dose prediction model, and calculating a predicted dose of sirolimus, wherein the predicted dose is daily average dose in a stable treatment period. The application develops a dosing dose prediction model aiming at three clinical problems of dose error, high risk group dosing risk and passive decision mode caused by development dynamics in the dosing of sirolimus for children.

Inventors

  • WANG XIAOLING
  • MAO HUAWEI
  • XU XIAOLIN
  • MAO XUETING
  • CHENG XIAOLING
  • LIU BO

Assignees

  • 首都医科大学附属北京儿童医院

Dates

Publication Date
20260508
Application Date
20260129

Claims (10)

  1. 1. A method of predicting a pediatric sirolimus dosing regimen, the method comprising: obtaining predicted parameter information of a sample to be detected, wherein the predicted parameters comprise weight, age, platelet count and creatinine; inputting the predicted parameter information into a dose prediction model, and calculating a predicted dose of sirolimus, wherein the predicted dose is a daily average dose in a stable treatment period; The construction method of the dose prediction model comprises the following steps: And iteratively training the machine learning model based on the training set sample comprising the prediction parameter data set, and taking the model obtained by training as the dose prediction model when the model obtained by training reaches the training stop condition.
  2. 2. The method of claim 1, wherein the age is 14 years or less.
  3. 3. The method of claim 1, wherein the predictive parameter information further comprises at least one of disease type, albumin, and low density lipoprotein cholesterol; Alternatively, the disease types include vascular malformations and primary immunodeficiency.
  4. 4. The method of claim 1, wherein the predictive parameter information further comprises at least one of aspartic acid aminotransferase, alanine aminotransferase, and gender.
  5. 5. The method of claim 1, wherein the predicted parameter information further comprises at least one of hemoglobin content, plasma concentration, urea, erythrocyte count, total bilirubin.
  6. 6. The method of claim 1, wherein the machine learning model comprises at least one of Catboost, DT, FT transducer, LGBM, linear Regression, resNet, ridge, SVM, tabPFN, XGBoost, preferably TabPFN.
  7. 7. The method for predicting the dosage of sirolimus to children according to claim 1, the method is characterized in that the prediction parameter data set is obtained by the following steps: Acquiring a total data set of training set samples including demographic characteristics, laboratory test indexes and treatment related indexes, and total sirolimus dose of the corresponding samples, wherein the total sirolimus dose is daily average dose in a stable treatment period; screening the total data set to obtain the prediction parameter data set; the demographic characteristics include age, sex, and weight, the laboratory test indicators include blood routine, liver and kidney function, and lipid profile, and the treatment-related indicators include blood concentration and disease type.
  8. 8. A computer device, characterized in that the device comprises a memory and a processor, the memory being for storing a computer program, the processor executing the computer program to carry out the steps of the method according to any one of claims 1-7.
  9. 9. A computer-readable storage medium, on which a computer program is stored, which computer program, when being executed by a processor, carries out the steps of the method according to any one of claims 1-7.
  10. 10. A computer program product comprising a computer program, characterized in that the computer program, when being executed by a processor, realizes the steps of the method according to any one of claims 1-7.

Description

Method, device, medium and program product for predicting sirolimus dosing of children Technical Field The present invention relates to the field of intelligent medicine, and more particularly, to a method, apparatus, medium and program product for predicting the dosage of sirolimus administered to children. Background Sirolimus (Sirolimus) is an inhibitor of mammalian target of rapamycin (mTOR), and is important in the treatment of pediatric organ transplant rejection, autoimmune diseases, and vascular malformations. However, its narrow therapeutic window (target blood concentration 5-15 ng/mL) and significant individual differences, combined with unique pharmacokinetic properties of the childhood population (e.g., differences in liver enzyme activity, rapid changes in body composition) make accurate dose regulation a clinically significant challenge. The current dosage regimen mainly depends on weight adjustment and empirical titration, and the first administration standard reaching rate is less than 40%, so that repeated blood concentration monitoring (TDM) and dosage adjustment period are prolonged, and treatment risks and medical burdens are increased. The study proves that the machine learning model can improve the dose prediction precision of narrow therapeutic window medicaments such as subliming Falin, tacrolimus and the like by integrating demographics, laboratory indexes and genetic factors. However, predictive model studies on pediatric sirolimus remain blank, with existing models multi-focusing on adult populations and generally ignoring physiological variables specific to children (such as growth indicators) and data high-dimensional characteristics. More importantly, the traditional model (such as XGBoost) is prone to overfitting when processing small samples of medical data, and the prediction accuracy is poor when the model is applied to the pharmaceutical field. Disclosure of Invention The present invention aims to solve at least one of the technical problems existing in the prior art. The invention provides a method, equipment, medium and program product for predicting the dosage of sirolimus administration of children, which aims at three clinical problems of dosage error caused by development dynamics, high risk group administration risk and passive decision mode in the process of the dosage of sirolimus administration of children, and develops a high-precision dosage prediction model. The application discloses a method for predicting sirolimus administration dosage of children in a first aspect, which comprises the following steps: obtaining predicted parameter information of a sample to be detected, wherein the predicted parameters comprise weight, age, platelet count PLT and creatinine CREA; inputting the predicted parameter information into a dose prediction model, and calculating a predicted dose of sirolimus, wherein the predicted dose is a daily average dose in a stable treatment period; The method for constructing the dose prediction model comprises the steps of obtaining a prediction parameter data set of a training set sample, carrying out iterative training on a machine learning model based on the training set sample comprising the prediction parameter data set, and taking the model obtained by training as the dose prediction model when the model obtained by training reaches a training stopping condition. In some embodiments, the age is less than or equal to 14 years old. In some embodiments, the predictive parameter information further includes at least one of a disease type, albumin ALB, and low density lipoprotein cholesterol LDL-C; Alternatively, the disease types include vascular malformations and primary immunodeficiency. In some embodiments, the predicted parameter information further comprises at least one of aspartic acid aminotransferase AST, alanine aminotransferase ALT, and Gender Gender. In some embodiments, the predicted parameter information further comprises at least one of hemoglobin content HGB, plasma concentration TDM, UREA UREA, red blood cell count RBC, total bilirubin TBIL. In some embodiments, the machine learning model includes at least one of Catboost, DT, FT transducer, LGBM, linear Regression, resNet, ridge, SVM, tabPFN, XGBoost, and preferably TabPFN. In some embodiments, the prediction parameter data set is obtained by: Acquiring a total data set of training set samples including demographic characteristics, laboratory test indexes and treatment related indexes, and total sirolimus dose of the corresponding samples, wherein the total sirolimus dose is daily average dose in a stable treatment period; screening the total data set to obtain the prediction parameter data set; The demographic characteristics include age, sex, and weight, the laboratory test indicators include blood routine, liver and kidney function, and lipid profile, and the treatment-related indicators include blood concentration TDM and disease type Source. In a second aspect the