GPT-4 for automated sequence-level determination of MRI protocols based on radiology request forms from clinical routine
Question Long MRI acquisition times limit patient access, making accurate protocol selection crucial for efficient diagnostics, though it's time-consuming and error-prone, especially for inexperienced residents. Findings GPT-4 generated MRI protocols of remarkable yet inconsistent quality, performing on par with an experienced resident in standardized fields, but moderately in musculoskeletal examinations. Clinical relevance The large language model can assist less experienced radiologists in determining detailed MRI protocols and counteract increasing workloads. The model could function as a semi-automatic tool, generating MRI protocols for radiologists' confirmation, optimizing resource allocation, and improving diagnostics and cost-effectiveness.
Authors
- Carsten Gietzen (ORCID: https://orcid.org/0000-0002-2354-3847)
- Lenhard Pennig (ORCID: https://orcid.org/0000-0002-6606-9313)
- Robert Hahnfeldt (ORCID: https://orcid.org/0000-0001-7997-3216)
- Andra-Iza Iuga (ORCID: https://orcid.org/0000-0002-3694-0235)
- Kenan Kaya (ORCID: https://orcid.org/0000-0003-3276-2571)
- David Maintz (ORCID: https://orcid.org/0000-0002-8942-3776)
- Robert Terzis (ORCID: https://orcid.org/0009-0007-1068-8477)
- Jan Paul Janßen (ORCID: https://orcid.org/0000-0003-0980-4606)
- Simon Lennartz (ORCID: https://orcid.org/0000-0002-3254-4809)
- Jonathan Kottlors (ORCID: https://orcid.org/0000-0002-0475-1472)
- Thomas Schömig (ORCID: https://orcid.org/0009-0001-7560-4285)
- Thomas Dratsch (ORCID: https://orcid.org/0000-0003-4014-7763)
- Lukas Müller (ORCID: https://orcid.org/0000-0002-0743-3357)
- Cansin Gözdas
Institutions
- University of Cologne (DE)
- Johannes Gutenberg University Mainz (DE)
- University Medical Center of the Johannes Gutenberg University Mainz (DE)
- University Hospital Cologne (DE)
Publication Details
- Journal
- European Radiology
- Published
- 2025-08-08
- DOI
- https://doi.org/10.1007/s00330-025-11888-4
- Citations
- 4
- Primary Topic
- Radiology practices and education
- Type
- article
- Field-Weighted Citation Impact
- 5.44
Funders
- Universität zu Köln