Facilitate Observational Studies of Alzheimer's Disease and Alzheimer's Disease-Related Dementias Using Ontology and Natural Language Processing

Yale University, New Haven CT

Investigators

Abstract

Project Summary As the 6th leading cause of death in the US, Alzheimer's disease (AD) and Alzheimer's disease-related dementias (ADRD) affect about 5.7 million Americans. However, up until now, our understanding of risk factors of AD/ADRD is still limited and our efforts on developing effective treatments for AD/ADRD have been greatly disappointing. Therefore, there is an urgent need to develop new methods to conduct AD/ADRD research more efficiently. One of the potential approaches is to leverage large, longitudinal, observational clinical data accumulated in electronic health records (EHRs). Nevertheless, current uses of EHRs for AD/ADRD research is very limited, often requiring manual data extraction and normalization (i.e., manual chart review), which is labor-intensive and time-consuming. Therefore, in this study, we plan to develop novel ontology and natural language processing (NLP) based informatics methods and tools to automatically extract and normalize AD/ADRD-related clinical data in EHRs, thus facilitating efficient AD/ADRD observational studies using EHRs. We propose the following three specific aims to achieve this goal: 1) Build an information model for EHR-based AD/ADRD research using a formal ontology representation approach; and 2) Extract and normalize AD/ADRD information in clinical documents using NLP technologies; and 3) Evaluate developed informatics methods and tools through demonstration studies and disseminate them to support observational AD/ADRD research.