Explore two innovative applications of Persistent Homology (PH) in oncology research during this 47-minute talk by Bernadette Stolz. Discover how PH combined with Machine Learning can predict leukemia relapse risk in pediatric patients, outperforming conventional methods. Learn about novel relational PH techniques using Dowker and Witness complexes to analyze spatial relationships in multispecies tumor microenvironment data. Gain insights into extracting biological information from topological features, including immune cell phenotypes and data-generating model parameters, which can significantly impact patient prognosis and treatment strategies.