Role of Model Size and Prompting Strategies in Extracting Labels from Free-Text Radiology Reports with Open-Source Large Language Models
Khosravi B, Dapamede T, Li F, Chisango Z, Bikmal A, Garg S, Owosela B, Khosravi A, Chavoshi M, Trivedi H, Wyles C, Purkayastha S, Erickson B, Gichoya J. Role of Model Size and Prompting Strategies in Extracting Labels from Free-Text Radiology Reports with Open-Source Large Language Models. Journal Of Imaging Informatics In Medicine 2025, 1-10. PMID: 40325326, DOI: 10.1007/s10278-025-01505-7.Peer-Reviewed Original ResearchMIMIC-CXRModel sizeImage classifierHuman annotatorsLanguage modelTraining medical image analysis modelsExtract labelsMedical image analysis modelsFree-text radiology reportsPre-trained modelsTrain image classifiersImage analysis modelsFracture detectionLabel noiseClassifier performanceRadiology reportsLabeling schemaOpen-sourceFree textLarge modelsClassifierExtraction accuracyRib fracture detectionDownstream performanceAnnotation
This site is protected by hCaptcha and its Privacy Policy and Terms of Service apply