Predicting emergency mortality risk in traumatic brain injury: comparative analysis of machine learning and large language model GPT-5
Proper threshold calibration enables LLMs to approximate ML accuracy while offering rapid deployment and interpretability. Prompt engineering combined with adaptive cut-off tuning may enhance clinical usability of LLM-based prediction systems.
Authors
- Kuan-Chi Tu
- Yushan Ma (ORCID: https://orcid.org/0000-0001-9933-6177)
- Ching-Lung Kuo
- Mei-I Sung
- Tee-Tau Eric Nyam
- Yung-De Kuo
- Chung-Feng Liu
Institutions
- National Sun Yat-sen University (TW)
- Chi Mei Medical Center (TW)
Publication Details
- Journal
- International Journal of Medical Informatics
- Published
- 2026-01-05
- DOI
- https://doi.org/10.1016/j.ijmedinf.2026.106268
- Citations
- 2
- Primary Topic
- Traumatic Brain Injury and Neurovascular Disturbances
- Type
- article
- Field-Weighted Citation Impact
- 78.30