---
title: "MedFiTRG: Jointly Learning Dynamic Temporal and Cross-Patient Graphs for Clinical Outcome Prediction"
authors:
  - "Shivani Gupta"
  - "Hari Om Kumar"
  - "Joydeep Chandra"
doi: "10.24963/ijcai.2026/746"
url: "https://ijcai.org/proceedings/2026/746"
pdf: "https://ijcai.org/proceedings/2026/0746.pdf"
venue: "Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence"
year: 2026
pages: "6707-6715"
track: "AI and Health"
keywords:
  - "Health data mining: Health data mining"
  - "Multimodal data: Multimodal data"
  - "Explainable AI: Explainable AI"
  - "Medical diagnosis: Medical diagnosis"
  - "Clinical decision support system: Clinical decision support system"
---

# MedFiTRG: Jointly Learning Dynamic Temporal and Cross-Patient Graphs for Clinical Outcome Prediction

Shivani Gupta, Hari Om Kumar, Joydeep Chandra

*Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence*, pages 6707-6715. DOI: [10.24963/ijcai.2026/746](https://doi.org/10.24963/ijcai.2026/746)

## Abstract

Integrating heterogeneous clinical modalities, structured electronic health records (EHRs), clinical text, and medical imaging is crucial for reliable clinical prediction, yet real-world data are often sparse and imbalanced. Furthermore, prior approaches treat temporal dynamics and inter-patient relationships in isolation, overlooking the dynamic interaction of patient trajectories across populations. We introduce a modality-enhanced dynamic temporal relational graph (MedFiTRG), a unified framework that jointly models sparsity, temporal dynamics, and cross-patient relational dependencies. MedFiTRG leverages modulated graph neural networks (MGNN) to learn modality-aware embeddings, enabling meaningful representation of sparse modalities through adaptive feature modulation. These embeddings are integrated into a temporal relational graph (TRG), where directed intra-patient edges capture longitudinal progression and dynamic inter-patient edges model population-level similarities for synchronized temporal-relational reasoning.
Extensive experiments on large-scale real-world datasets across four clinical tasks demonstrate that MedFiTRG achieves superior or comparable performance against state-of-the-art baselines, improving Macro-F1 from 0.155 to 0.310 for length of stay (LOS) classification and achieving an AUROC of 0.939 for mortality prediction (↑6.45%).
The code is available at https://anonymous.4open.science/r/MedFiTRG-2714