Construction of prediction models for hypertensive nephropathy based on machine learning
10.3969/j.issn.1673-9701.2025.15.002
- VernacularTitle:基于机器学习构建高血压肾病预测模型研究
- Author:
Mingming LIU
1
;
Hong WANG
;
Zhecheng WANG
;
Dan CHEN
Author Information
1. 台州市第一人民医院中医科,浙江 台州 318020
- Publication Type:Journal Article
- Keywords:
Machine learning;
Hypertension;
Hypertensive nephropathy;
Prediction model;
Logistic regression;
Support vector machine;
Artificial neural network
- From:
China Modern Doctor
2025;63(15):7-10,110
- CountryChina
- Language:Chinese
-
Abstract:
Objective To explore the construction of a prediction model for hypertensive nephropathy(HN)based on machine learning.Methods A total of 318 hypertensive patients who visited Taizhou First People's Hospital from April 2023 to March 2024 were included and divided into a training set and a validation set at a ratio of 7:3.Least absolute shrinkage and selection operator(LASSO)algorithm was used to select clinical features from the training set,and 12 clinically significant variables were obtained from 18 clinical variables.Based on the Python 3.10 programming language,the training set was used to train the model.Taking the 12 clinically significant indicators were used as input variables,and whether the occurrence of HN was used as the outcome variable.Three machine learning algorithms,namely logistic regression,support vector machine,and artificial neural network,were used to construct prediction models.The test set was used for internal validation of three models.The performance of the models was compared through accuracy,area under the receiver operating characteristic curve,recall rate,precision,and F1.Results Among 12 clinically significant variables screened by the LASSO algorithm,cystatin C and urine protein qualitative were found to be the most predictive.The accuracy,area under the receiver operating characteristic curve,recall rate,precision,and F1 values of the Logistic regression,support vector machine,and artificial neural network prediction models constructed by machine learning was 0.94,0.96,0.95,0.87,0.91;0.94,0.97,0.96,0.86,0.91;0.91,0.94,0.93,0.80,0.86,respectively.Conclusion Logistic regression,support vector machine,and artificial neural network based on machine learning all have good predictive effects on the progression of hypertensive patients to HN.Among them,the predictive effects of Logistic regression and support vector machine are similar and better than artificial neural network prediction model.
