Background

Machine learning (ML) offers unique insights into clinical data. In this study, we evaluated the impact of alimentary and biliopancreatic limb lengths in Roux-en-Y Gastric Bypass (RYGB) surgery on weight loss. We conducted a retrospective analysis on 88 patients who underwent RYGB surgery from 2020-2022 using propensity score matching for baseline weights. Using k-means clustering, patients were stratified into four groups based on limb lengths: Roux limbs above and below 137 cm, and biliopancreatic limbs above and below 90 cm. We assessed changes in weight and BMI at 3, 6, and 12 months post-surgery. Although all groups had substantial BMI reductions, there were no significant differences between groups with differing limb lengths. Both longer and shorter Roux limb groups showed similar BMI reductions (-13.6 kg/m² vs. -14.9 kg/m² respectively). Biliopancreatic limb groups demonstrated comparable results, with BMI changes of -14.9 kg/m² and -14 kg/m², in longer limbs versus shorter limbs. All groups had p-values >0.05. Heatmap analysis indicated a weak negative correlation between limb lengths and outcomes. The study concludes that neither Roux nor biliopancreatic limb lengths significantly influence weight loss at 3, 6, and 12 months post-RYGB surgery. This finding aligns with existing literature and underscores the need for further research. The application of ML techniques offers a promising direction for future studies. ML utilization in larger cohorts may enable the development of predictive models to optimize surgical procedures for individual patients, enhancing postoperative outcomes.