CLOZE: Letting AI Turn Doctors’ Notes into Smarter Medical Maps
Medical ontologies are like maps of diseases, drugs, and symptoms. They power search, decision support, and research—but they miss many real-world terms used in doctors’ notes.
A new framework called CLOZE taps large language models to read clinical notes and suggest new concepts and where they fit in the hierarchy—without extra training data.
- Zero-shot: Works out of the box; no labeled data needed.
- Privacy-first: Automatically removes protected health information (PHI).
- Richer coverage: Finds disease-related entities and their parent/child links.
- Scalable & cost-efficient: Automates what used to be slow expert curation.
In tests, CLOZE extended ontologies accurately and at scale, pointing to better tools for biomedical research and clinical informatics.
Paper: https://arxiv.org/abs/2511.16548v1
Paper: https://arxiv.org/abs/2511.16548v1
Register: https://www.AiFeta.com
#AI #Healthcare #ClinicalAI #LLM #NLP #MedicalInformatics #Privacy #Ontology
