titleTRACE: Tracking Multimodal Clinical Trajectories for Inpatient Risk Prediction
authorsMinjun Kim, Jong Hak Moon
date2026-10-19
venuearXiv (preprint, not peer reviewed)
idarXiv:2610.25765
upvotes2 (Hugging Face Daily Papers, 2026-11-01)
linksabs pdf hf
abstractAccurate inpatient risk prediction requires modeling heterogeneous clinical evidence that accumulates irregularly over time. This challenge spans in-hospital mortality, prolonged length of stay, and readmission, each with task-specific observation windows. Existing multimodal approaches combine structured health record data, chest x-rays, and clinical text, but primarily focus on modality fusion or missing-modality compensation, with limited attention to temporal trajectories over an admission. We propose TRACE (TRAjectory modeling of multimodal Clinical Events), a daily-scale framework organizing clinical events into day-level trajectories, separating day-invariant global context from day-specific local events drawn from EHR data, chest x-rays, radiology reports, and discharge summaries. The trajectory is encoded using temporal attention, allowing unavailable modalities to emit no tokens. A phenotype-guided contrastive objective further organizes admissions under heterogeneous disease distributions. Across cohorts from MIMIC-IV, eICU, and MIMIC-III comprising nearly 180,000 total stays, we evaluate TRACE on mortality, length of stay, and readmission prediction. TRACE achieves the highest AUROC across all nine dataset-task settings and the highest ACC in eight, outperforming baselines. Ablation studies validate our attention masks and contrastive learning, while trajectory-level attribution localizes risk-associated hospital days and modalities.

cite

@misc{kim2026trace,
  title={TRACE: Tracking Multimodal Clinical Trajectories for Inpatient Risk Prediction},
  author={Minjun Kim and Jong Hak Moon},
  year={2026},
  eprint={2610.25765},
  archivePrefix={arXiv},
  url={https://arxiv.org/abs/2610.25765}
}