Can generative artificial intelligence provide accurate medical advice?: a case of ChatGPT versus Congress of Neurological Surgeons management of acute cervical spine and spinal cord injuries clinical guidelines
ChatGPT-4 was moderately accurate when generating recommendations that aligned with the clinical guidelines. The model frequently aligned with low evidence and therapeutic recommendations but exhibited inferior performance on topics that contained high-quality evidence or pertained to diagnostic and clinical assessment strategies. Medical practitioners should monitor its usage until further models can be rigorously trained on medical data.
Authors
- Alexander Yu (ORCID: https://orcid.org/0000-0002-7246-2269)
- Fady Y. Hijji (ORCID: https://orcid.org/0000-0001-9820-3911)
- Bashar Zaidat
- Jonathan Markowitz (ORCID: https://orcid.org/0000-0002-0987-0761)
- Michael Saturno (ORCID: https://orcid.org/0000-0002-1132-0662)
- Samuel Kang-Wook Cho (ORCID: https://orcid.org/0000-0001-7511-2486)
- Akiro H. Duey (ORCID: https://orcid.org/0000-0003-0308-0512)
- Jun Kim (ORCID: https://orcid.org/0000-0002-6114-2673)
- Mateo Restrepo Mejia (ORCID: https://orcid.org/0009-0003-0457-3308)
- Wasil Ahmed
Institutions
- Icahn School of Medicine at Mount Sinai (US)
Publication Details
- Journal
- Asian Spine Journal
- Published
- 2025-03-04
- DOI
- https://doi.org/10.31616/asj.2024.0301
- Primary Topic
- Artificial Intelligence in Healthcare and Education
- Type
- article
- Field-Weighted Citation Impact
- 0.00