Abstract IA025: AI-powered discovery of biomarkers from histopathological images in ovarian cancer

Abstract High-grade serous ovarian cancer (HGSOC) presents a significant clinical challenge due to limited prognostic tools, hindering personalized treatment strategies. Improved risk stratification through accurate survival prediction can lead to tailored interventions and better patient care. Artificial intelligence (AI), particularly deep learning, offers a promising avenue for extracting and integrating the rich information contained within haematoxylin and eosin (H&E)-stained whole-slide images (WSIs) to discover diagnostic and prognostic biomarkers. In this talk, I will present work from my group on the development of AI and deep learning approaches for predicting patient outcomes from histopathological images, including genomic alterations and treatment response. I will describe our explainability frameworks, designed to uncover mechanisms of disease progression and enhance the trustworthiness of AI models. I will also highlight challenges in this field, drawing on insights from our recent work on the AI in Histopathology Explorer ( https://www.histopathexpo.ai/ ). Citation Format: Heba Sailem. AI-powered discovery of biomarkers from histopathological images in ovarian cancer [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Advances in Ovarian Cancer Research; 2025 Sep 19-21; Denver, CO. Philadelphia (PA): AACR; Cancer Res 2025;85(18_Suppl):Abstract nr IA025.