From Trajectories to Phenotypes: Disease Progression as Structural Priors for Multi-organ Imaging Representation Learning

TL;DR

This paper introduces a trajectory-aware distillation framework that leverages disease trajectory models to enhance multi-organ imaging representations, improving disease prediction accuracy.

Contribution

It presents a novel method that transfers structural knowledge from disease trajectory transformers to imaging encoders, improving disease prediction in imaging data.

Findings

Trajectory-aware pretraining improves disease discrimination and time-to-onset prediction.

Embedding space similarities in IDP and trajectory models partially align.

The approach enhances robustness of imaging-based disease prediction under cohort constraints.

Abstract

Imaging-derived phenotypes (IDPs) summarize multi-organ physiology but provide only static snapshots of diseases that evolve over time. In contrast, longitudinal electronic health records encode disease trajectories through temporal dependencies among past diagnosis events and comorbidity structure. We hypothesize that IDPs and disease trajectories contain partially shared disease-relevant structure. We propose a trajectory-aware distillation framework that transfers structural knowledge from a generative disease trajectory Transformer into an organ-wise IDP encoder. A population-scale trajectory model trained on longitudinal diagnosis sequences produces subject-level embeddings that supervise IDP representation learning via geometry-preserving alignment. During downstream prediction, trajectory and imaging representations can also be fused via cross-attention. Across 159 diseases in the UK Biobank cohort, trajectory-aware pretraining consistently improves both discrimination (AUC) and time-to-onset prediction (MAE), with the largest gains for low-prevalence diseases. Similarity relationships in IDP embedding space also align with those in trajectory space, providing supportive evidence for partially aligned representation geometry. These results suggest that population-scale generative disease models can serve as structural priors for data-limited imaging modalities, improving robustness under realistic cohort constraints.