Natural language processing to phenotype coronary computed tomography angiography: Development, validation, and initial results of a large multi-institution cohort

The role and uptake of coronary computed tomography angiography (CTA) in the evaluation and management of cardiovascular disease continues to expand rapidly throughout the United States and worldwide.1 Importantly, as the amount of cardiovascular imaging data grows exponentially, there is an unmet need to develop efficient techniques to accurately extract and characterize such information for research, quality improvement, and population health efforts. Natural language processing (NLP) is one such technology that has been developed across a variety of clinical domains, working to transform clinical reports into structured data available for large-scale analytic work.
Source: Journal of Cardiovascular Computed Tomography - Category: Radiology Authors: Tags: Correspondence Source Type: research