UdonCare: Hierarchy Pruning for Unseen Domain Discovery in Predictive Healthcare
This addresses the problem of enabling model generalization to unseen domains in healthcare for providers, but it is incremental as it builds on domain generalization with a novel application to clinical settings.
The paper tackled the challenge of domain generalization in clinical prediction models without explicit domain labels by proposing UdonCare, a hierarchy-guided method that prunes medical ontologies to identify latent domains and decompose domain-invariant information, achieving superiority over eight baselines on MIMIC-III and MIMIC-IV datasets across four clinical prediction tasks.
Healthcare providers often divide patient populations into cohorts based on shared clinical factors, such as medical history, to deliver personalized healthcare services. This idea has also been adopted in clinical prediction models, where it presents a vital challenge: capturing both global and cohort-specific patterns while enabling model generalization to unseen domains. Addressing this challenge falls under the scope of domain generalization (DG). However, conventional DG approaches often struggle in clinical settings due to the absence of explicit domain labels and the inherent gap in medical knowledge. To address this, we propose UdonCare, a hierarchy-guided method that iteratively divides patients into latent domains and decomposes domain-invariant (label) information from patient data. Our method identifies patient domains by pruning medical ontologies (e.g. ICD-9-CM hierarchy). On two public datasets, MIMIC-III and MIMIC-IV, UdonCare shows superiority over eight baselines across four clinical prediction tasks with substantial domain gaps, highlighting the untapped potential of medical knowledge in guiding clinical domain generalization problems.