From Records to Relationships: An Ontology-Based Knowledge Graph Framework for Cancer Data Interoperability.

Healthcare data's rapid growth and complexity reveal a significant challenge: we lack effective approaches to integrate multimodal information into unified, interpretable structures. Existing methods typically focus either on ontology mapping or knowledge graph construction, but they don't fully address interoperability, reasoning, and explainability together. This paper presents a foundation for ontology-driven cancer data analysis that brings together knowledge graphs (KGs), SNOMED CT, and DICOM imaging metadata. Our methodology aligns ontologies and applies reasoning to semantically integrate structured clinical data, unstructured text, and imaging sources, which generates interpretable, patient-centered insights, and is illustrated through a compact colorectal cancer cohort identification scenario. We designed this structure to address the shortcomings of current approaches and deliver explainable, interoperable healthcare analytics for precision oncology.