Search papers, labs, and topics across Lattice.
This paper reviews the concept of medical world models, which adapt AI's world-modeling techniques to healthcare by creating internal simulators that represent patient-state dynamics and clinical interventions. It emphasizes the need for a shift from static diagnostic outputs to dynamic models that can predict disease progression and inform tailored treatment strategies. The authors propose a structured roadmap for integrating capabilities in patient-state construction, clinical dynamics modeling, and intervention decision support to enhance the utility of medical AI systems.
Static labels in medical AI are giving way to dynamic models that can simulate disease evolution and guide personalized interventions.
Medical diagnosis and treatment are dynamic processes in which patient states evolve over time and clinical interventions alter future outcomes. Although current medical AI can detect disease, estimate risk and generate reports, many systems still return static labels or scores, offering limited insight into how illness may progress or how alternative interventions may reshape its trajectory. Medical world models adapt the world-model idea from artificial intelligence to healthcare by learning internal simulators of patient-state dynamics. Their long-term goal is to help clinicians anticipate deterioration, compare treatment-conditioned futures and tailor care to individual patients. Yet relevant work remains scattered across foundation models, longitudinal modelling, disease simulation, treatment-effect estimation, reinforcement learning and digital twins. To bridge this gap, this review outlines a roadmap for advancing medical AI from isolated diagnosis and prediction toward medical world models that simulate disease evolution and support intervention decisions. This roadmap is organized around three coupled capabilities: patient-state construction, clinical dynamics modelling and intervention decision support. Across representative systems, the comparison highlights what each capability contributes and how partial components can be integrated into more mature perception--dynamics--planning systems. Finally, we identify the challenges involved in turning plausible rollouts into clinically useful simulators. Related literature is available at https://github.com/1999kevin/awesome_medical_world_models.
