Paper: A repository of semantic types in the MIMIC II database clinical notes

ACL ID W14-3414
Title A repository of semantic types in the MIMIC II database clinical notes
Venue Proceedings of the BioNLP Shared Task 2013 Workshop
Session
Year 2014
Authors

The MIMIC II database contains 1,237,686 clinical documents of vari- ous kinds. A common task for researchers working with this database is to run MetaMap, which uses the UMLS Metathesaurus, on those documents to identify specific semantic types of entities mentioned in them. However, this task is computationally expensive and time-consuming. Research in many groups could be accelerated if there were a community-accessible set of outputs from running MetaMap on this document collection, cached and available on the MIMIC-II website. This paper describes a repository of all MetaMap output from the MIMIC II database, publicly available, assuming compliance with usage agree- ments required by UMLS and MIMIC-II. Additionally, software for manipulating MetaMap output, available on Source- Forge ...