PS55: Using AI/ML to Integrate Real-World Evidence and Patient-Reported Outcomes in Drug Development for Hematologic Malignancy– Methods and Applications

In today's digital era, hematologic malignancies-such as lymphoma, multiple myeloma, and acute myeloid leukemia-demand advanced analytical approaches that synthesize diverse data sources. These include real-world evidence (RWE), patient-reported outcomes (PROs), and multimodal data from electronic health records, genomics, imaging, and clinical endpoints. This session will explore how artificial intelligence and machine learning (AI/ML) enable seamless data integration to enhance prognostic modeling, optimize clinical trial design, and accelerate patient-centered drug development in hematology/oncology. Presentations will highlight innovative AI/ML methodologies-such as deep neural networks, Bayesian frameworks, and causal inference techniques-applied to real-world datasets, large prospective registries, and PROs capturing symptoms, quality of life, and treatment tolerability. Key discussion areas include: • Integrating complex multimodal data for robust predictive modeling • Validating surrogate endpoints and improving prognostic tools • Designing efficient, adaptive clinical trials informed by AI/ML insights • Addressing challenges such as algorithmic bias and regulatory considerations The session aims to foster cross-functional dialogue among statisticians, data scientists, clinicians, and regulators, emphasizing strategies to ensure transparency, mitigate bias, and prioritize meaningful patient outcomes.