AI-Enhanced Bariatric and Endobariatric Surgery: Predictive Analytics, Complication Surveillance and Intelligent Patient Care

Artificial intelligence (AI) and machine learning (ML) are increasingly influencing bariatric surgery by enabling personalized care and better risk management. This review summarizes recent evidence on AI/ML in surgical and endoscopic weight-loss procedures, based on approximately 300 studies from 2018 to 2025. Findings cover preoperative risk prediction, postoperative monitoring, intraoperative AI, endobariatric applications, long-term outcome prediction, and ethical considerations. Preoperative AI models, such as neural networks and deep learning, modestly improved complication prediction with AUROC gains of 0.02–0.12 over traditional methods. Predictive models for sleeve gastroplasty achieved high accuracy for 12-month weight loss using early postoperative data. Long-term models reliably forecast five-year BMI, and deep learning outperformed scores in predicting diabetes remission. However, limitations include class imbalance, rare-adverse-event detection, limited external validation, calibration gaps, and scant evidence for some endobariatric procedures. Intraoperative computer vision is still in its early stages of development. Successful AI implementation requires multicenter validation, explainability, adherence to standards, and versatile strategies for diverse healthcare settings.