Accelerating Medical Record Data Abstraction and Analysis in Muscular Dystrophy
LLMs promise to accelerate EMR abstraction for rare diseases such as muscular dystrophy, but F1 scores for LLMs currently lag manual abstractions for unstructured data. Llama 3-8b demonstrated superior performance to the 4 other models tested. Metadata such as ICD code counts may help prioritize high-yield cases for surveillance and research purposes.
Authors
- Yara Mohamed (ORCID: https://orcid.org/0009-0005-3835-9443)
- Jiatan Huang
- Huixue Zhou (ORCID: https://orcid.org/0000-0002-6524-5506)
- Erjia Cui (ORCID: https://orcid.org/0000-0003-3576-2892)
- Peter B. Kang (ORCID: https://orcid.org/0000-0002-4270-7325)
- Carla D. Zingariello (ORCID: https://orcid.org/0000-0001-9884-2190)
- Geetanjali Rajamani (ORCID: https://orcid.org/0000-0002-1769-4307)
- Magali Jorand‐Fletcher (ORCID: https://orcid.org/0009-0006-7369-6696)
- Rui Zhang (ORCID: https://orcid.org/0000-0001-8258-3585)
- Kody A. DeGolier
- Annette Xenopoulos-Oddsson
Institutions
- University of Minnesota (US)
- Muscular Dystrophy Association (US)
- Allen Institute for Brain Science (US)
- University of Florida (US)
- University of Minnesota Medical Center (US)
- Minnesota Department of Health (US)
- Florida College (US)
Publication Details
- Journal
- Neurology Clinical Practice
- Published
- 2025-09-23
- DOI
- https://doi.org/10.1212/cpj.0000000000200542
- Primary Topic
- Genomics and Rare Diseases
- Type
- article
- Field-Weighted Citation Impact
- 0.00