Multimodal Disease Progression Modeling via Spatiotemporal Disentanglement and Multiscale Alignment
This work addresses the challenge of integrating multimodal data for disease progression modeling, which is incremental as it builds on existing methods by introducing specific disentanglement and alignment techniques.
The paper tackled the problem of modeling disease progression from multimodal longitudinal data by addressing redundancy in chest X-ray sequences and temporal misalignment with EHR data, resulting in state-of-the-art performance on disease progression identification and ICU prediction tasks as demonstrated on the MIMIC dataset.
Longitudinal multimodal data, including electronic health records (EHR) and sequential chest X-rays (CXRs), is critical for modeling disease progression, yet remains underutilized due to two key challenges: (1) redundancy in consecutive CXR sequences, where static anatomical regions dominate over clinically-meaningful dynamics, and (2) temporal misalignment between sparse, irregular imaging and continuous EHR data. We introduce $\texttt{DiPro}$, a novel framework that addresses these challenges through region-aware disentanglement and multi-timescale alignment. First, we disentangle static (anatomy) and dynamic (pathology progression) features in sequential CXRs, prioritizing disease-relevant changes. Second, we hierarchically align these static and dynamic CXR features with asynchronous EHR data via local (pairwise interval-level) and global (full-sequence) synchronization to model coherent progression pathways. Extensive experiments on the MIMIC dataset demonstrate that $\texttt{DiPro}$ could effectively extract temporal clinical dynamics and achieve state-of-the-art performance on both disease progression identification and general ICU prediction tasks.