. Multimodal AI for Atypical Hyperplasia Diagnosis: A Comparative Study of CNNs, GNNs, and Hybrid Models. In 21st European Congress on Digital Pathology. Barcelona: The European Society of Digital and Integrative Pathology; 2025. 
(592.84 KB)
(592.84 KB) Abstract
The diagnosis of atypical hyperplasia in endometrial aspirates is complex due to morphological variations across the menstrual cycle, often leading to interobserver variability. This study explores the use of artificial intelligence to enhance diagnostic accuracy by developing an AI-assisted framework to improve reproducibility and clinical decision-making in gynecological pathology.
