Overview

Case Search enables semantic search across 167,000+ published patient cases. You describe the clinical situation in plain language, and Asclevor returns the most relevant published cases.

How a query flows through the system

Natural-language query
        ↓
     Embedding
        ↓
Vector similarity search
        ↓
     HNSW index
        ↓
Most relevant patient cases

Because search is semantic rather than keyword-based, queries match cases that describe the same clinical picture in different words.

What you get back

Each result includes:

• Patient case text — the case description itself
• PMID — the PubMed identifier of the source article
• Similarity score — how semantically close the case is to your query
• Metadata — title, journal, and publication year

See the full response schema in the API reference.