JoVE - Journal of Visualized Experiments
JoVE Visualize
Contact Us

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

  • 0Department of Signal Theory and Communications, and Telematics Engineering, University of Valladolid, Paseo de Belén, 15, 47011 Valladolid, Spain.
Nutrients +

|

April 26, 2025

|

April 26, 2025

View abstract on PubMed

Summary

This summary is machine-generated.

Machine learning, especially Support Vector Machines (SVMs), can predict successful weight loss after Sleeve bariatric surgery. Key predictors include biochemical markers, anthropometric data, and psychological factors.

Area Of Science

  • Bariatric Surgery
  • Machine Learning in Healthcare
  • Obesity Treatment Outcomes

Background

  • Obesity is a critical global health challenge.
  • Bariatric surgery is the most effective treatment for severe obesity.
  • Predicting postoperative weight loss is crucial due to outcome variability.

Purpose Of The Study

  • To identify variables that predict successful weight loss one year post-Sleeve bariatric surgery.
  • Define success as exceeding 30% weight loss.
  • Evaluate machine learning models for predictive accuracy.

Main Methods

  • Utilized a dataset of 94 patients from 2013-2018.
  • Applied machine learning algorithms: Random Forest, Multilayer Perceptron, XGBoost, Decision Tree, Logistic Regression, and Support Vector Machines (SVMs).

Main Results

  • The SVM model achieved the highest performance with 88% accuracy and an AUC of 0.76.
  • Identified key predictive variables: potassium, folic acid, alkaline phosphatase, height, transferrin, weight, BMI, triglycerides, Beck Depression Test score, and insulin.

Conclusions

  • Machine learning models, particularly SVMs, show promise in predicting Sleeve bariatric surgery success.
  • Successful weight loss is multifactorial, influenced by biochemical, anthropometric, and psychological factors.

Keywords: