Multi-Modal AI Models in Healthcare: Toward Integrated Clinical Reasoning and Disease Prediction

This session is part of the AI in Healthcare Preconference Forum and additional registration is required.

The future of clinical decision-making lies at the intersection of multi-modal data and AI-powered reasoning. In this interview-style session, Okan Ekinci, MD (CMIO, Roche Information Solutions), speaks with Peter McCaffrey, MD (Chief AI Officer of UTMB Galveston) about the emerging frontier of multi-modal AI models and their potential to transform healthcare delivery.

Together, they will explore how diverse data streams—such as lab values, imaging, pathology, genomics, and unstructured clinical notes—can be integrated to build comprehensive models that not only predict disease but also emulate clinician-like reasoning.

The conversation will touch on the practical challenges and strategic opportunities of deploying such systems in large academic medical centers, the role of AI innovation hubs like UTMB's, and how partnerships with industry and diagnostic leaders can accelerate adoption.

This session moves beyond theory to focus on the practical building blocks of trustworthy AI. Attendees will learn about scaling challenges, the critical value of local model validation, and the infrastructure needed for integrated diagnostic intelligence, providing a practical roadmap for digital health leaders, data scientists, clinicians, and executives.