Experimental image-based AI models are very good at extracting new information from standard H&E pathology slides. However, the biological interpretation of these novel findings is a challenge. In this webinar, Dr. Anna Laury, Clinical Researcher at the University of Helsinki, will explain how they used scanned H&E slides of high-grade serous carcinoma of the tube/ovary (HGSC) together with spatially resolved transcriptomics to explore whether AI findings actually represent biological differences in these tumors.
High-grade serous carcinoma of the tube/ovary (HGSC) is characterized by aggressive behavior, chemotherapy resistance, and low 5-year survival, but also exhibits striking variability in outcome. Our understanding of this disease is limited, partly due to considerable tumor heterogeneity, and predicting any individual patient’s response to therapy is challenging.
Dr. Laury’s team used Aiforia® Create to train an AI model to predict patient outcome in HGSC by identifying tumor regions that are highly associated with outcome status. The tumor regions identified by the AI model are currently indistinguishable by pathologists. They then used spatially resolved transcriptomics to profile these AI-identified tumor regions and identify molecular features related to disease outcome.
Watch the webinar to hear more about this proof-of-concept study, which shows that AI-guided spatial transcriptomic analysis improves recognition of biologic features relevant to patient outcomes.
