Assessing the performance of ChatGPT and Bard/Gemini against radiologists for Prostate Imaging-Reporting and Data System classification based on prostate multiparametric MRI text reports
This study highlights the limitations of LLMs in accurately classifying PI-RADS categories from clinical text reports. While the performance of LLMs has improved with newer versions, caution is warranted before integrating such technologies into clinical practice.
Authors
- Dimitri A. Kessler (ORCID: https://orcid.org/0000-0002-1813-1039)
- Tristan Barrett (ORCID: https://orcid.org/0000-0002-1180-1474)
- Nadeem Shaida (ORCID: https://orcid.org/0000-0002-6843-7458)
- Kang‐Lung Lee (ORCID: https://orcid.org/0000-0002-1648-7154)
- Yu-Lun Kuo (ORCID: https://orcid.org/0000-0002-8131-0734)
- Iztok Caglič (ORCID: https://orcid.org/0000-0002-3432-3540)
Institutions
- National Yang Ming Chiao Tung University (TW)
- University of Cambridge (GB)
- Cambridge University Hospitals NHS Foundation Trust (GB)
- Taipei Veterans General Hospital (TW)
- Bridge University (SS)
- Addenbrooke's Hospital (GB)
Publication Details
- Journal
- British Journal of Radiology
- Published
- 2024-11-13
- DOI
- https://doi.org/10.1093/bjr/tqae236
- Citations
- 15
- Primary Topic
- Artificial Intelligence in Healthcare and Education
- Type
- article
- Field-Weighted Citation Impact
- 1.80
Funders
- Cancer Research UK
- National Institute for Health and Care Research
- Department of Health and Social Care
- Engineering and Physical Sciences Research Council
- NIHR Cambridge Biomedical Research Centre