Development and Validation of an Artificial Intelligence-Based Model to Predict Gastroesophageal Reflux Disease After Sleeve Gastrectomy

• DOI: 10.1007/s11695-022-06112-x
• Publication Date: 2022-05-21T08:02:46Z
• Authors: Sameh Hany Emile; Waleed Ghareeb; Hossam Elfeki; Mohamed El Sorogy; Amgad Fouad; Mohamed Elrefai
• Keywords: Sleeve gastrectomy; Cut-off

Prediction of the onset de novo gastroesophageal reflux disease (GERD) after sleeve gastrectomy (SG) would be helpful in decision-making and selection optimal bariatric procedure for every patient. The present study aimed to develop an artificial intelligence (AI)-based model predict GERD SG help clinicians surgeons decision-making.A prospectively maintained database patients with severe obesity who underwent was used development AI using all available data points. dataset arbitrarily split into two parts: 70% training 30% testing. Then ranking variables performed steps. Different learning algorithms were used, best that showed maximum performance selected further steps machine learning. A multitask platform determine cutoff points top numerical predictors GERD.In total, 441 (76.2% female) a mean age 43.7 ± 10 years included. ensemble outperformed other models. achieved AUC 0.93 (95%CI 0.88-0.99), sensitivity 79.2% (95% CI 57.9-92.9%), specificity 86.1% 70.5-95.3%). five ranked age, weight, preoperative GERD, size orogastric tube, distance first stapler firing from pylorus.An AI-based prediction developed. had excellent accuracy, yet moderate specificity. Further prospective multicenter trials are needed externally validate