Document Type
Conference Proceeding
Publication Date
10-2015
Abstract
ezDI uses large and extensive knowledge graph to enhance linguistics, NLP and ML techniques to improve structured data extraction from millions of EMR records. It then normalizes it, and maps it with various computer-processable nomenclature such as SNOMED-CT, RxNorm, ICD-9, ICD-10, CPT, and LOINC. Furthermore, it applies advanced reasoning that exploited domain-specific and hierarchical relationships among entities in the knowledge graph to make the data actionable. These capabilities are part of its highly scalable AWS deployed heath intelligence platform that support healthcare informatics applications, including Computer Assisted Coding (CAC), Computerized Document Improvement (CDI), compliance and audit, and core measures and utilization, as well as support improved decision making that involve identification of patients at risk, patterns in diseases, outcome prediction, etc. This paper focuses on the key role of its semantic approach and techniques.
Repository Citation
Goswami, R.,
Shah, N.,
& Sheth, A. P.
(2015). ezDI's Semantics-Enhanced Linguistic, NLP, and ML Approach for Health Informatics. .
https://corescholar.libraries.wright.edu/knoesis/1078
Included in
Bioinformatics Commons, Communication Technology and New Media Commons, Databases and Information Systems Commons, OS and Networks Commons, Science and Technology Studies Commons
Comments
Presented at the 14th International Semantic Web Conference, Bethlehem, PA, October 11-15, 2015.