Scholarlink

The valvuloplasty procedure includes inserting possible balloon catheters over the narrowed valve, typically over an extra-stiff guide wire, in order to create valvotomy. The balloons are then inflated with diluted contrast material. Such type of intervention requires a variables length of stay for patients, strongly depending on several factors. The scope of the paper is to analyze the length of stay (LOS) for 25 patients that underwent valvuloplasty by using Multiple Linear Regression and predict LOS by Artificial Intelligence models. The best prediction was obtained using Decision Trees (DT) and Gradient Boosted Trees (GBT) algorithms with an accuracy that exceeds 80%. Even the linear multiple regression model was able to characterize the phenomenon well with an R2 greater than 0.5.

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

Advanced Data Analysis Techniques to Study Valvuloplasty Hospitalization: A Multicenter Study

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Marta Rosaria Marino,
  • Anna Borrelli,
  • Fabiana Rubba,
  • Maria Triassi,
  • Giovanni Improta

摘要

The valvuloplasty procedure includes inserting possible balloon catheters over the narrowed valve, typically over an extra-stiff guide wire, in order to create valvotomy. The balloons are then inflated with diluted contrast material. Such type of intervention requires a variables length of stay for patients, strongly depending on several factors. The scope of the paper is to analyze the length of stay (LOS) for 25 patients that underwent valvuloplasty by using Multiple Linear Regression and predict LOS by Artificial Intelligence models. The best prediction was obtained using Decision Trees (DT) and Gradient Boosted Trees (GBT) algorithms with an accuracy that exceeds 80%. Even the linear multiple regression model was able to characterize the phenomenon well with an R2 greater than 0.5.