Explainable Framework for Ontology-Based Similarity : A Use Case on SNOMED CT

Article Article scientifique Article Explainable Framework for Ontology-Based Similarity : A Use Case on SNOMED CT eng 2026-05-21 author Baudin Alexis Division of Medical Information Sciences, Geneva University Hospitals, Geneva, Switzerland.,Department of Radiology and Medical Informatics, University of Geneva, Geneva, Switzerland.,Geneva Digital Health Hub, Faculty of medicine, University of Geneva, Geneva, Switzerland. 426484 Département de radiologie et informatique médicale (RADIO) author Gaudet-Blavignac Christophe 0000-0001-6527-5898 Division of Medical Information Sciences, Geneva University Hospitals, Geneva, Switzerland.,Department of Radiology and Medical Informatics, University of Geneva, Geneva, Switzerland. 645533 author Lovis Christian 0000-0002-2681-8076 Division of Medical Information Sciences, Geneva University Hospitals, Geneva, Switzerland.,Department of Radiology and Medical Informatics, University of Geneva, Geneva, Switzerland. 29016 Département de radiologie et informatique médicale (RADIO) author Bjelogrlic Mina 0000-0002-6922-3283 Division of Medical Information Sciences, Geneva University Hospitals, Geneva, Switzerland.,Department of Radiology and Medical Informatics, University of Geneva, Geneva, Switzerland. 918489 Département de radiologie et informatique médicale (RADIO) 443-447 10.3233/SHTI260194 42174870 0926-9630 In healthcare, being able to efficiently manipulate and compare concepts in ontologies is crucial to enable semantic interoperability of clinical data. Most ontology-based similarity functions return a single score with little actionable justification, limiting their trust and use in clinical workflows. In this paper, we focus on the explainability of similarity computations rather than on defining yet another similarity measure. We introduce an explainable similarity framework that returns a score together with a small set of named pivots and readable paths from concept to pivot. The contribution of each pivot is a combination of intrinsic topological specificity and path-length proximity, aggregated into a similarity score. Qualitative cases derived from SNOMED CT show specific pivots for clinically close pairs and broad pivots for distant ones. An internal AUC of 0.998 matches classical measures while providing path-based explanations; explainability metrics confirm shorter paths, deeper pivots, and higher pivot impact for close pairs. SNOMED CT Semantic similarity Clinical ontologies Explainability Interpretable method Knowledge graph Systematized Nomenclature of Medicine Semantics Electronic Health Records Natural Language Processing Humans Dewey 616.0757 62 Département de radiologie et informatique médicale 227 false 344 Interfaces Homme-machine en milieu clinique 610 Studies in health technology and informatics 336 243db994-3e52-4bb3-87a2-fcbaf19b52c2 Article (Published version) Public SHTI-336-SHTI260194.pdf 344657 application/pdf CC BY-NC-4.0 12b5ed02a97eb0502502d21406261815ccf09e33b9e6b70382be217de28a5874 openAccess Faculté de médecine / Section de médecine clinique / Département de radiologie et informatique médicale

Title	Published in	Access level	OA Policy	Year	Views	Downloads

Explainable Framework for Ontology-Based Similarity : A Use Case on SNOMED CT	Studies in health technology and informatics			2026	4	42

Archive ouverte UNIGE

Baudin, Alexis