A 2025 study from BMC Gastroenterology reveals that an AI system using endoscopic ultrasound effectively differentiates small gastric tumors. With the ResNet50 model, subtle imaging features are classified with high accuracy, offering promise in early diagnosis and treatment planning. This advancement may enhance clinical decision-making in gastroenterology.

Researchers Caixia Fang et al. have developed a LightGBM-based predictive model that accurately identifies cognitive impairment in cholestasis patients. The model integrates factors like age and plasma D-dimer levels for early risk stratification, offering a promising tool for precision medicine. (Source: BMC Gastroenterology, 2025)

A 2025 case-control study conducted by researchers at Guangxi Medical University presents a machine learning model, particularly using a Random Forest, to predict liver cancer risk in chronic hepatitis B. Key markers such as AST/ALT, BLR, and AFP stand out. Consider monitoring these parameters to facilitate timely clinical interventions.