STM-GNN : Space-Time-and-Memory Graph Neural Networks for Predicting Multi-Drug Resistance Risks in Dynamic Patient Networks

Conférence Chapitre d'actes STM-GNN : Space-Time-and-Memory Graph Neural Networks for Predicting Multi-Drug Resistance Risks in Dynamic Patient Networks eng 2025-06-23 2025 author Geissbuhler Damien 885252 author Bornet Alban 0000-0002-7266-627X 402515 Département de radiologie et informatique médicale (RADIO) author Marques Catarina author Anjos André author Pereira Sónia author Teodoro Douglas 0000-0001-6238-4503 682380 Département de radiologie et informatique médicale (RADIO) Douglas Henrique Teodoro 160-169 10.1007/978-3-031-95838-0_16 978-3-031-95837-3 https://link.springer.com/10.1007/978-3-031-95838-0_16 <p>Hospital-acquired infections (HAIs), particularly those caused by multidrug-resistant (MDR) bacteria, pose significant risks to vulnerable patients. Accurate predictive models are important for assessing infection dynamics and informing infection prediction and control (IPC) programmes. Graph-based methods, including graph neural networks (GNNs), offer a powerful approach to model complex relationships between patients and environments but often struggle with data sparsity, irregularity, and heterogeneity. We propose the space-time-and-memory (STM)-GNN, a temporal GNN enhanced with recurrent connectivity designed to capture spatiotemporal infection dynamics. STM-GNN effectively integrates sparse, heterogeneous data combining network information from patient-environment interactions and internal memory from historical colonization and contact patterns. Using a unique IPC dataset containing clinical and environmental colonization information collected from a long-term healthcare unit, we show that STM-GNN effectively addresses the challenges of limited and irregular data in an MDR prediction task. Our model reaches 0.84 AUROC, and achieves the most balanced performance overall compared to classic machine learning algorithms, as well as temporal GNN approaches.</p> Hospital acquired infection Temporal graph neural network Dewey 616.0757 Springer Cham Lecture Notes in Computer Science 15734 62 Département de radiologie et informatique médicale 227 false 1379 DS4DH - Data Science for Digital Health 1035 Artificial Intelligence in Medicine : 23rd International Conference, AIME 2025, Pavia, Italy, June 23–26, 2025, Proceedings, Part I Bellazzi, R., Juarez Herrero, J.M., Sacchi, L. & Zupan, B. June 23–26, 2025 Pavia (Italy) 60f80e26-15e6-400e-8e64-8d3283574c7c Proceedings chapter (Published version) Restricted STM-GNN-_Space-Time-and-Memory_Grap.pdf 2959169 application/pdf b6ff680d1be5981b9e5343e0847f66bf475782f01d9268ca48b3e63efa6f2f94 f6a626f5-c4f4-4cb7-b053-c961c112862f Proceedings chapter (Accepted version) Public Temporal_graph_networks_for_infecti.pdf 4692600 application/pdf Restricted 2027-01-01 026f206b34c08949e2a17957aca144014221d3deaeea11ed3b66d0d4eb8be72b embargoedAccess Faculté de médecine / Section de médecine clinique / Département de radiologie et informatique médicale

Title	Published in	Access level	OA Policy	Year	Views	Downloads

STM-GNN : Space-Time-and-Memory Graph Neural Networks for Predicting Multi-Drug Resistance Risks in Dynamic Patient Networks	Artificial Intelligence in Medicine : 23rd International Conference, AIME 2025, Pavia, Italy, June 23–26, 2025, Proceedings, Part I			2025	21	0

Archive ouverte UNIGE

Geissbuhler, Damien