Differentiating between GPT-generated and human-written feedback for radiology residents
Currently, GPT-3.5 cannot match human experts in delivering specific, nuanced feedback for radiology residents. Compared to humans, GPT-3.5 also performs worse in distinguishing between actual and synthetic comments. These insights could guide the development of more sophisticated algorithms to produce higher-quality feedback, supporting faculty development.
Authors
- Andrew D. Chung (ORCID: https://orcid.org/0000-0001-5114-4379)
- Benjamin Y. M. Kwan (ORCID: https://orcid.org/0000-0002-1124-3206)
- Zier Zhou (ORCID: https://orcid.org/0000-0002-4890-2842)
- Arsalan Rizwan
- Nick Rogoza
Institutions
- Queen's University (CA)
- Logikos (United States) (US)
Publication Details
- Journal
- Current Problems in Diagnostic Radiology
- Published
- 2025-02-18
- DOI
- https://doi.org/10.1067/j.cpradiol.2025.02.002
- Citations
- 5
- Primary Topic
- Radiology practices and education
- Type
- article
- Field-Weighted Citation Impact
- 6.80
Funders
- Association Canadienne des Radiologistes