Response accuracy of GPT-4 across languages: insights from an expert-level diagnostic radiology examination in Japan
GPT-4 exhibits higher accuracy when responding to English-translated questions compared to original Japanese questions, a trend not observed with German or Chinese translations. Accuracy improves with higher-quality English translations, underscoring the importance of high-quality translations in improving GPT-4's response accuracy to diagnostic radiology questions in non-English languages and aiding non-native English speakers in obtaining accurate answers from large language models.
Authors
- Jin Yamamura (ORCID: https://orcid.org/0000-0003-3990-5125)
- Yoshitaka Toyama (ORCID: https://orcid.org/0000-0003-0027-9681)
- Mirei Abe (ORCID: https://orcid.org/0009-0009-0487-7313)
- Ayaka Harigai (ORCID: https://orcid.org/0000-0001-5732-0716)
- Kei Takase (ORCID: https://orcid.org/0000-0003-0931-9942)
- Masahiro Kawabata (ORCID: https://orcid.org/0009-0003-3362-3458)
- Li Li (ORCID: https://orcid.org/0000-0003-2542-2392)
- Mitsutoshi Nagano
Institutions
- Universität Hamburg (DE)
- Tohoku University (JP)
- Tohoku University Hospital (JP)
- University Medical Center Hamburg-Eppendorf (DE)
- Tohoku Medical and Pharmaceutical University Hospital (JP)
Publication Details
- Journal
- Japanese Journal of Radiology
- Published
- 2024-10-28
- DOI
- https://doi.org/10.1007/s11604-024-01673-6
- Citations
- 9
- Primary Topic
- Artificial Intelligence in Healthcare and Education
- Type
- article
- Field-Weighted Citation Impact
- 1.08