Implementing public health analytical services: Grid enabling of MetaMap

Abstract

Public health data could be used to assist with public health surveillance and decision support. However, in most cases data has to be transformed into a coded format to make it computable and amiable to quasi real time analytical processing. Natural language processing (NLP) systems, which aim to accurately extract and encode biomedical information in a standard format, have a great potential in surveillance. NLP methods are complex, difficult, and expensive to implement. Its implementation, in most cases, is well beyond the technical expertise and resources available in Public Health organizations. Making NLP systems available as a service can greatly improve access to this methodology by public health officials and potentially enhance disease surveillance. MetaMap is a comprehensive biomedical NLP system, and has been shown to perform well for numerous applications. We describe how we have implemented MetaMap as a grid service to make it available to the public health community.