Predicting Weight Loss Success After Gastric Sleeve Surgery: A Machine Learning-Based Approach DOI Open Access

Manuel Fernández Domínguez,

Isabel Herrera Montano, Juan José López Gómez

et al.

Nutrients, Journal Year: 2025, Volume and Issue: 17(8), P. 1391 - 1391

Published: April 21, 2025

Background/Objectives: Obesity is a global health issue, and in this context, bariatric surgery considered the most effective treatment for severe cases. However, postoperative outcomes vary widely among individuals, driving development of tools to predict body weight loss success. The main objective paper evaluate predictive variables successful one year after Sleeve surgery, defining success as exceeding 30%. Methods: A dataset 94 cases was included study. Data were collected between 2013 2018 from Nutrition Section Endocrinology Department Eastern Area Valladolid, Spain. Machine learning algorithms applied Random Forest, Multilayer Perceptron, XGBoost, Decision Tree, Logistic Regression, Support Vector Machines (SVMs). Results: SVM model demonstrated best performance, attaining an accuracy 88% area under curve (AUC) 0.76 with 95% CI 0.5238 0.9658. identified potassium (K), folic acid, alkaline phosphatase (ALP), height, transferrin, weight, mass index (BMI), triglyceride (Tg), Beck Depression Test score, insulin levels. Conclusions: In conclusion, study highlights potential machine models, particularly (SVMs), predicting surgery. key include biochemical markers, anthropometric measures, psychological factors, emphasizing multifactorial nature outcomes.

Language: Английский

Predicting Weight Loss Success After Gastric Sleeve Surgery: A Machine Learning-Based Approach DOI Open Access

Manuel Fernández Domínguez,

Isabel Herrera Montano, Juan José López Gómez

et al.

Nutrients, Journal Year: 2025, Volume and Issue: 17(8), P. 1391 - 1391

Published: April 21, 2025

Background/Objectives: Obesity is a global health issue, and in this context, bariatric surgery considered the most effective treatment for severe cases. However, postoperative outcomes vary widely among individuals, driving development of tools to predict body weight loss success. The main objective paper evaluate predictive variables successful one year after Sleeve surgery, defining success as exceeding 30%. Methods: A dataset 94 cases was included study. Data were collected between 2013 2018 from Nutrition Section Endocrinology Department Eastern Area Valladolid, Spain. Machine learning algorithms applied Random Forest, Multilayer Perceptron, XGBoost, Decision Tree, Logistic Regression, Support Vector Machines (SVMs). Results: SVM model demonstrated best performance, attaining an accuracy 88% area under curve (AUC) 0.76 with 95% CI 0.5238 0.9658. identified potassium (K), folic acid, alkaline phosphatase (ALP), height, transferrin, weight, mass index (BMI), triglyceride (Tg), Beck Depression Test score, insulin levels. Conclusions: In conclusion, study highlights potential machine models, particularly (SVMs), predicting surgery. key include biochemical markers, anthropometric measures, psychological factors, emphasizing multifactorial nature outcomes.

Language: Английский

Citations

0