Artificial Intelligence Models for the Diagnosis of Gastrointestinal Disorders: A Systematic Review of Literature (#270)

Authors

Llamo-Sánchez, Johan Iván

Esquén-Salazar, Erenia Vanessa

Dios-Castillo, Christian Abraham

Abstract

This study examines Artificial Intelligence (AI) models' effectiveness in detecting and classifying gastrointestinal disorders (GID) based on complex patterns and biometric data. The research highlights the impact of different AI approaches, focusing on Deep Learning (DL), Machine Learning (ML), and hybrid ML+DL models. The results show that CNN-based DL models perform exceptionally well when handling large volumes of data, achieving high accuracy, especially in identifying conditions such as polyps, ulcers, and Crohn's disease. Hybrid models that combine ML and DL architectures offer superior performance, with lower variability in results and higher diagnostic accuracy.