Session (CTI)

Purpose:This workshop will introduce AI and ML approaches—including unsupervised and supervised ML, LLMs, digital twins, and RL—to address unmet needs in HEOR. The session will help attendees understand when next-generation AI adds value beyond traditional analytics; how these approaches can strengthen personalized therapy in RWE studies; and review methodological, ethical, and transparency considerations for responsible HEOR application. Description:AI innovations—including generative and multimodal LLMs, digital twins, and RL—are reshaping HEOR by enabling deeper insights from real-world data, identifying meaningful patient subgroups, and informing individualized treatment strategies. This workshop provides an accessible overview of these emerging methods, their applications, and considerations for ethical implementation. The session will feature presentations: •Dr. Marcum (moderator): Overview of the evolution of ML/AI in HEOR, highlighting unmet needs. •Dr. Tian: Case study showing how integrating unsupervised and supervised ML uncovers meaningful patient subgroups in schizophrenia and inform personalized treatment strategies. •Dr. Peng: Application of digital twins and multimodal LLMs to predict treatment responses in oncology and mental health disorders, emphasizing potential to enhance individualized decision support. •Dr. Lo-Ciganic: Use of advanced ML to characterize complex treatment pathways using sunburst plots to reveal real-world prescribing patterns. RL Illustrations show support for providers tailoring treatment strategies for patients with opioid use disorder and co-occurring mental health conditions, along with key ethical considerations like fairness, bias detection, and transparency. Interactive components include live polling and scenario-based questions to engage attendees in evaluating strengths, limitations, and implementation challenges. The workshop will conclude with a moderated discussion on best practices for integrating next-generation AI/ML into HEOR to support personalized therapy and evidence-informed decision making.