PubMed HealthSearch

PubMed · 6843410

Computerized speech recognition for anesthesia recordkeeping.

Abstract

EARS, a computerized speech recognition system, has been developed to simplify data entry for automated anesthesia recordkeeping. As an isolated-word, speaker-dependent system, it uses a predefined vocabulary of about 350 words, each of which must first be pronounced by the user to train the computer to his voice. Thereafter, data are entered by spoken words separated by brief pauses and organized into sentences according to a specified syntax. Completed sentences are read back to the user by a speech synthesizer to verify the recognition. Errors can be corrected before they are recorded. Attention to human factors is an important design goal. Performance has not yet been measured, but appears acceptable.

Explore related subjects

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

A J Sarnat. Computerized speech recognition for anesthesia recordkeeping.. https://pubmed.ncbi.nlm.nih.gov/6843410/

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

KEEP EXPLORING

Related citations