Machine Learning Model in Obesity to Predict Weight Loss One Year after Bariatric Surgery: A Pilot Study.

Area of Science:

• Bariatric Surgery
• Machine Learning in Healthcare
• Obesity Management

Background:

• Roux-en-Y gastric bypass (RYGB) is a key treatment for severe obesity.
• Insufficient total weight loss (TWL) is a common challenge post-RYGB.
• Factors influencing RYGB success require further elucidation.

Purpose of the Study:

• To assess the feasibility and reliability of machine learning (ML) for predicting weight loss after RYGB.
• To identify clinical, anthropometric, and biochemical predictors of poor weight loss response.
• To aid in selecting patients for bariatric surgery.

Main Methods:

• Retrospective analysis of 118 patients undergoing RYGB.
• Application of ML techniques, including local linear embedding (LLE).
• Utilized evolutionary algorithms for model optimization and parameter adjustment.

Main Results:

• Identified key variables associated with one-year %TWL: obstructive sleep apnea, osteoarthritis, insulin treatment, preoperative weight, insulin resistance index, apolipoprotein A, uric acid, complement component 3, and vitamin B12.
• The ML model achieved 71.4% accuracy in classifying patients with <30% TWL.
• Moderate discriminatory precision was observed in the validation set.

Conclusions:

• Machine learning models show potential in assisting patient selection for bariatric surgery.
• ML can help identify individuals at risk of insufficient weight loss post-RYGB.
• This approach supports personalized treatment strategies in severe obesity management.