A GPT‐4o‐Powered Framework for Identifying Cognitive Impairment Stages in Electronic Health Records
Our GPT-4o-powered framework achieves high performance in CI classification, outperforming traditional embedding models and demonstrating potential for automated chart review. However, clinical deployment requires careful consideration of misclassification risks, making a human-in-the-loop approach essential for reliability and safety. This hybrid model underscores AI's role in dementia diagnosis while ensuring interpretability and risk mitigation in real-world applications.
Authors
- Colin Magdamo (ORCID: https://orcid.org/0000-0001-8965-4630)
- Samad Amini (ORCID: https://orcid.org/0000-0002-2063-6220)
- John R. Dickson (ORCID: https://orcid.org/0000-0003-0135-7928)
- Shibani S. Mukerji (ORCID: https://orcid.org/0000-0002-5677-6954)
- Ioannis Ch. Paschalidis (ORCID: https://orcid.org/0000-0002-3343-2913)
- Yu Leng (ORCID: https://orcid.org/0000-0002-7039-4127)
- Deborah Blacker
- Ana‐Maria Vranceanu
- Yignan He
- Sudeshna Das
- Lidia Maria V. Moura
- Christine S Ritchie
- M. Brandon Westover
Institutions
- Beth Israel Deaconess Hospital (US)
- Massachusetts General Hospital (US)
Publication Details
- Journal
- Alzheimer s & Dementia
- Published
- 2025-12-01
- DOI
- https://doi.org/10.1002/alz70857_107622
- Primary Topic
- Machine Learning in Healthcare
- Type
- article
- Field-Weighted Citation Impact
- 0.00