Search PubMed⌕ Search
Home/Archive/PubMed

PubMed · 8563286

Experience with a mixed semantic/syntactic parser.

P J Haug·S Koehler·L M Lau·P Wang·R Rocha·S M Huff
View original source ↗Export metadata ↓

Abstract

The value of the computerized medical record is derived in part from the availability of medical information in a coded form accessible to manipulation by processes designed for automated decision support, medical research, and computer assistance in the management of health care delivery. To meet these needs medical reports captured and stored as natural language documents must be encoded. Below we discuss an ongoing formative process aimed at developing a natural language understanding system for chest x-ray reports. Comparative data showing the progress of this process is presented.

Explore related subjects

Hospital Information SystemsHumansLanguageMedical Records Systems, ComputerizedNatural Language ProcessingRadiography, ThoracicSemanticsSoftwareJournal ArticleResearch Support, U.S. Gov't, P.H.S.

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

P J Haug, S Koehler, L M Lau, P Wang, R Rocha, S M Huff. 1995. Experience with a mixed semantic/syntactic parser.. https://pubmed.ncbi.nlm.nih.gov/8563286/

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related citations

PubMed1997

Effect of computerized requisition of radiology examinations on the transmission of clinical information.

RATIONALE AND OBJECTIVES: The authors developed a method to evaluate the availability and accuracy of clinical data proffered by clinicians when ordering radiologic examinations with a computer order-entry system. METHODS: Two thousand consecutive clinical indications for a spectrum of pediatric imaging studies were scrutinized for accuracy by means of computerized chart review, verbal communication with clinical attending staff, and reference to surgical and laboratory results. The indications were classified as appropriate, incorrect, misleading, or incomplete. RESULTS: Of the 2,000 stated indications, 1,464 (73%) provided a reasonable, if minimal, amount of clinical information; however, in 376 (19%) cases the diagnosis or proximate indication was incorrect, in 108 (5%) cases the data were incomplete, and in 52 (3%) cases the information was misleading. CONCLUSION: Computer order-entry systems can improve the transmission of clinical information but they allow misinformation to be provided.

Hospital Information Systems↗