An open-source fine-tuned large language model for radiological impression generation: a multi-reader performance study.

Journal: BMC medical imaging
PMID: 39333958

Abstract

BACKGROUND: The impression section integrates key findings of a radiology report but can be subjective and variable. We sought to fine-tune and evaluate an open-source Large Language Model (LLM) in automatically generating impressions from the remainder of a radiology report across different imaging modalities and hospitals.

Authors

  • Adrian Serapio
    Department of Radiology and Biomedical Imaging, University of California, San Francisco, San Francisco, CA, USA.
  • Gunvant Chaudhari
    From the School of Medicine (J.V.C., G.C.) and Department of Radiology and Biomedical Imaging (C.P.H., O.A.G., L.P.S., A.M.R., Y.L.), University of California, San Francisco, 505 Parnassus Avenue, M-391, San Francisco, CA 94143-0628.
  • Cody Savage
    Department of Radiology, University of Maryland Medical Center, Baltimore, MD, USA.
  • Yoo Jin Lee
    Department of Internal Medicine.
  • Maya Vella
    Department of Radiology and Biomedical Imaging, University of California, San Francisco, San Francisco, CA, USA.
  • Shravan Sridhar
    Department of Radiology and Biomedical Imaging, University of California, San Francisco, San Francisco, CA, USA.
  • Jamie Lee Schroeder
    MedStar Georgetown University Hospital, Washington, DC, USA.
  • Jonathan Liu
    Dept. of Biomedical Engineering, Johns Hopkins University, Baltimore MD.
  • Adam Yala
    Department of Electrical Engineering and Computer Science, CSAIL, MIT, Cambridge, USA.
  • Jae Ho Sohn
    Radiology & Biomedical Imaging, UCSF Medical Center, 505 Parnassus Ave, San Francisco, CA, 94158, USA. sohn87@gmail.com.

Keywords

External Resources

Popular Topics
Recent Journals