extract-clinical-entities-to-fhir

To convert clinical text to FHIR resources, this Skill extracts clinical entities from de-identified text and maps accepted spans into deterministic FHIR R4 resources like Conditions and Observations. It applies OpenMed model inference and confidence filtering to build validated FHIR Bundles.

Yes, you can extract clinical entities offline without exposing PHI by processing synthetic or de-identified text locally. The Skill preserves safe audit offsets and maps results to FHIR R4 resources without inventing terminology codes or exposing protected health information.

The best way to map medical coding terminology to FHIR R4 Bundles is by applying approved terminology mappings during entity extraction. The Skill filters accepted clinical spans by label and confidence, then assembles deterministic FHIR resources with Bundle validation.

Yes, this clinical NLP workflow supports medication extraction and disease detection. It analyzes local clinical text with OpenMed models, filters results by label and confidence, and maps detected diseases and medications into appropriate FHIR Conditions and MedicationStatements.

A key limitation of clinical entity extraction for healthcare interoperability is that it requires synthetic or de-identified text input. The Skill does not invent terminology codes and mandates FHIR Bundle validation against receiver requirements to ensure safe audit handling.

Yes, you need OpenMed model inference to generate synthetic Patient resources. The Skill requires OpenMed models to analyze clinical text, filter detected entities by confidence and label, and map accepted spans into validated FHIR Bundles containing synthetic Patient and Condition resources.