Making data available to researcher through extracting essential information from oncological medical records is one of Optum’s key oncology data initiatives. To that end, Optum has developed a natural language processing system (NLP) to provide oncology- related insights by identifying desired oncological concepts and extracting information from the Optum electronic health records (HER) data asset. The information extraction process uses three approaches to extract relevant entities in the text and relationship:
- Entity extraction. The extraction of a concept or entity represented by lexical units or phrases in the free text.
- Relation extraction. The extraction of the relationships between entities.
- Frame extraction. The extraction of the logical semantic group of lexical units and the collection of any relevant relations.
Advantages of Optum’s NLP approach include scalability and comprehensive, consistent, and reliable extraction, leading to effective and highly accurate results. Extraction results for specific oncology concepts such as stage and TNM consistently exceed 90% precision. Find out more from the white paper here.
(Source: Optum Clinical natural language processing, August 12, 2020)