BI-RADS Category Assignments by GPT-3.5, GPT-4, and Google Bard: A Multilanguage Study
Andrea Cozzi, Katja Pinker, Andri Hidber, Tianyu Zhang, Luca Bonomo, Roberto Lo Gullo, Blake Christianson, Marco Curti, Stefania Rizzo, Filippo Del Grande, Ritse M. Mann, Simone Schiaffino
Abstract
Based on breast imaging reports written in three languages, GPT-3.5, GPT-4, and Google Bard achieved moderate agreement with human reader–assigned Breast Imaging Reporting and Data System categories but also yielded a high percentage of discordant category assignments that would negatively impact clinical management.
Topics & Concepts
MedicineMcNemar's testBI-RADSMammographyReferralBreast imagingBreast cancerRadiologyFamily medicineCancerInternal medicineStatisticsMathematicsArtificial Intelligence in Healthcare and EducationRadiology practices and educationAI in cancer detection