The absence of a flexible, robust, and accurate natural language interface is a significant barrier to the direct use of computer-based patient records by dental clinicians. While providing patient care, dentists, hygienists and assistants are handicapped in using a keyboard and mouse to interact with a computer, primarily because of infection control concerns. The objective of this proposal is to develop and evaluate a prototype dental charting system with a speech-driven interface that will allow the dentist to chart dental conditions using natural language. The system will use Natural Language Processing (NLP) to extract the key concepts associated with 16 dental conditions from transcribed dental examinations. These concepts, coded using the standardized terminologies, would provide a structured summary of a patient's initial dental exam. The proposal has two aims: 1) evaluate the accuracy of speech recognition technology for clinical dental examinations; and 2) develop and evaluate an NLP application for mapping transcribed text to a structured dental chart. This proposal describes a new, exploratory and innovative research project that could radically impact the practice of dental charting. Expected outcomes for this proposal include: 1) an understanding of the accuracy of speech recognition for real-time dictated dental exams; and 2) NLP-based tools to automatically chart restorative and periodontal conditions for each tooth into a structured dental chart. This developmental work will provide a strong foundation for developing a chairside NLP-based dental charting application that would automatically generate a structured dental chart suitable for chairside decision support. ? ? ?
| Irwin, Jeannie Y; Harkema, Henk; Christensen, Lee M et al. (2009) Methodology to develop and evaluate a semantic representation for NLP. AMIA Annu Symp Proc 2009:271-5 |
