Natural Language Processing

The effort proposed uses Natural Language Processing (NLP) and Knowledge Base technologies to address Data Entry Automation, E&M coding for billing, and Syndromic Surveillance use cases within the Electronic Medical Record (EMR) field. The approach is to leverage Language and Computing’s TeSSI NLP products and Carnegie Mellon University’s Scone symbolic knowledgebase technology. The transition plan is two fold. First, the goal is to improve the free text data entry user interface and calculate E&M codes from clinical documentation within AHLTA system by adapting and integrating TeSSI. Second, the plan is to adapt and integrate Scone to TeSSI and other medical, geographical, nation, transportation and news feeds to provide outbreak recognition capability to existing syndromic surveillance systems such as DoD ESSENCE and CDC Biosense. An Open Architecture will be applied to enable affordable technology insertion cycles by standardizing the interface between the NLP and KB, and subcomponents. This also allows multiple components to support different parts of the problem. For example, one NLP can be used for E&M coding for billing use case, and another for syndromic surveillance use case.