LLM-MINE: Large Language Model based Alzheimer's Disease and Related Dementias Phenotypes Mining from Clinical Notes

LLM-MINE is a proposed framework leveraging Large Language Models for the automatic extraction of Alzheimer's Disease and Related Dementias (ADRD) phenotypes from unstructured clinical notes. The framework demonstrated superior performance in phenotype clustering, achieving an Adjusted Rand Index (ARI) of 0.290 and Normalized Mutual Information (NMI) of 0.232, significantly surpassing traditional biomedical Named Entity Recognition (NER) and dictionary-based methods. This advancement is crucial for practitioners as it enhances the ability to mine clinically relevant signals from electronic health records, facilitating early detection and staging of ADRD.