RadSearch, a Semantic Search Model for Accurate Radiology Report Retrieval with Large Language Model Integration

2025/04/01 by Cody Savage, Gunvant Chaudhari, Andrew D. Smith +1 · 1 voice · 1 citation
Medicine · Computer Science · #COVID-19 diagnosis using AI #Topic Modeling #Artificial Intelligence in Healthcare and Education

paper · doi:10.1148/radiol.240686

openalex publication_date 2025/04/01 · openalex created_date 2025/10/10 · openalex updated_date 2026/06/11

Abstract

RadSearch, a specialized semantic search model for radiology reports trained using a scalable method, achieved state-of-the-art performance in retrieving relevant radiology reports for queried report findings, improving the diagnostic accuracy of a large language model.

Cited by

• Pre‐Imaging Clinical Factors Associated With Cardiac MR Image Quality Using Large Language Model‐Enabled Data Extraction

Discussions

• In a retrospective study, @ucsfci2.bsky.social's Dr. Jae Ho Sohn developed a scalable method for training semantic search models for radiology reports. ⬇️ [bsky, 0 points, 0 comments]