Rheumatic Digital Twin: A Proposed Machine Learning-Based Multi-Modal Framework to Inform Clinical Decision-Making (Preprint)

Medical digital twins (MDTs) apply advanced machine learning algorithms on longitudinally collected data to transform patient care. We introduce an MDT framework designed for rheumatic diseases, integrating multi-modal data from electronic health records. By designing an architecture that can effectively represent each data modality, both at baseline and during treatment, we aim to construct comprehensive and informative digital twins that encapsulate a patient’s individual characteristics, clinical history, and current health status, thereby supporting the implementation of precision medicine.