Scholarlink

<p>To develop and validate a machine learning model by combining blood markers with retinal structural parameters to predict the risk of primary glaucoma and provide a risk assessment tool for early detection of primary glaucoma. A case–control study was adopted, and 268 glaucoma patients and 1072 non-glaucoma patients were included in Sichuan Provincial People’s Hospital from February 2023 to November 2024. Candidate predictors include demographic data, blood markers, and retinal structural parameters. Five feature selection methods and six machine learning algorithms were used to develop the model and compare their predictive performance. Eleven potential predictors associated with primary glaucoma were screened and identified, which were subsequently used to construct a prediction model for primary glaucoma. The cross-validation results indicate that the model built using the Xtreme Gradient Boosting (XGBoost) algorithm (AUC: 0.986, 95% CI 0.974–0.995) showed relatively better comprehensive predictive performance among the six machine learning algorithms. The SHAP diagram visualized and interpreted the risk prediction model for primary glaucoma. The XGBoost model exhibited relatively better comprehensive predictive performance, and it is of great significance to apply these models to predict and screen for primary glaucoma early.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Prediction of primary glaucoma: development and validation of multiple machine learning models

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年4月6日

查看全文
  • Huiwang Zhang,
  • Zhenghong Luo,
  • Huixian Xiong,
  • Mingxuan Xie,
  • Ping Shuai

摘要

To develop and validate a machine learning model by combining blood markers with retinal structural parameters to predict the risk of primary glaucoma and provide a risk assessment tool for early detection of primary glaucoma. A case–control study was adopted, and 268 glaucoma patients and 1072 non-glaucoma patients were included in Sichuan Provincial People’s Hospital from February 2023 to November 2024. Candidate predictors include demographic data, blood markers, and retinal structural parameters. Five feature selection methods and six machine learning algorithms were used to develop the model and compare their predictive performance. Eleven potential predictors associated with primary glaucoma were screened and identified, which were subsequently used to construct a prediction model for primary glaucoma. The cross-validation results indicate that the model built using the Xtreme Gradient Boosting (XGBoost) algorithm (AUC: 0.986, 95% CI 0.974–0.995) showed relatively better comprehensive predictive performance among the six machine learning algorithms. The SHAP diagram visualized and interpreted the risk prediction model for primary glaucoma. The XGBoost model exhibited relatively better comprehensive predictive performance, and it is of great significance to apply these models to predict and screen for primary glaucoma early.