Using convolutional neural network to predict remission of diabetes after gastric bypass surgery: a machine learning study from the Scandinavian Obesity Surgery Register (Preprint)

BACKGROUND Prediction of diabetes remission is an important topic in the evaluation of patients with type-2 diabetes (T2D) before bariatric surgery. While several high-quality predictive indices are available, artificial intelligence (AI) algorithms offer the potential for higher predictive capability. OBJECTIVE The objective was to construct and validate an AI prediction model for diabetes remission after Roux-en-Y gastric bypass surgery. METHODS Patients who underwent surgery from 2007 until 2017 were included in the study, with collection of individual data from the Scandinavian Obesity Surgery Registry (SOReg), the Swedish National Patients Register, the Swedish Prescribed Drugs Register, and Statistics Sweden. A 7-layer convolution neural network (CNN) model was developed using 80% of patients randomly selected from SOReg and 20% of patients for external testing. The predictive capability of the CNN model and currently used scores (DiaRem, Ad-DiaRem, DiaBetter and IMS) were compared. RESULTS In total, 8057 patients with T2D were included in the study. At 2 years after surgery 77.1% achieved pharmacological remission, while 62.2% achieved complete remission. The area under the receiver operating curve (AUC) for the CNN-model for pharmacological remission was 0.85 [95% confidence interval (CI): 0.83-0.86] during validation, and 0.83 for the final test, which was 9-12% better than the traditional predictive indices. AUC for complete remission was 0.83 (95% CI: 0.81-0.85) during validation, and 0.82 for the final test, which was 9-11% better than the traditional predictive indices. CONCLUSIONS The CNN method had better predictive capability compared to traditional indices for diabetes remission. However, further validation is needed in other countries to evaluate its external generalizability.