Author: Jun Li, MS (2018)
Primary advisor: Hua Xu, PhD
Committee members: Cui Tao, PhD; Jiajie Zhang, PhD; Yang Gong, MD, PhD
PhD thesis, The University of Texas School of Health Information Sciences at Houston.
ABSTRACT
Extracting and encoding clinical information captured in unstructured clinical documents with standard medical terminologies is vital to enable secondary use of clinical data from practice. SNOMED CT is the most comprehensive medical ontology with broad types of concepts and detailed relationships and it has been widely used for many clinical applications. However, few studies have investigated the use of SNOMED CT in clinical information extraction.
In this dissertation research, we developed a fine-grained information model based on the SNOMED CT and built novel information extraction systems to recognize clinical entities and identify their relations, as well as to encode them to SNOMED CT concepts. Our evaluation shows that such ontology-based information extraction systems using SNOMED CT could achieve state-of-the-art performance, indicating its potential in clinical natural language processing.