N J Zemmel, S M Park, E J Maurer, L F Leslie, R F Edlich
{"title":"Evaluation of VoiceType Dictation for Windows for the radiologist.","authors":"N J Zemmel, S M Park, E J Maurer, L F Leslie, R F Edlich","doi":"","DOIUrl":null,"url":null,"abstract":"<p><p>The accuracy of the radiology vocabulary domain of the IBM VoiceType Dictation for Windows speech recognition system for use in a clinical setting was evaluated. Six men and one woman dictated a case report under several conditions. The impact of vocabulary domain, microphone position, personal enrollment, and background noise were assessed. The dictations were exported to Microsoft Word 6.0, printed, and graded for accuracy. The data were expressed as a percentage correct. The enrollment and training processes took an average of 5 hours to complete. The accuracy of VoiceType Dictation was significantly reduced by speech volume and background noise level so that it could not be used to dictate clinical information accurately in a hospital setting. Although IBM's VoiceType Dictation with the radiology vocabulary domain has improved speech recognition, it is not a reliable dictation system for accurate and efficient clinical dictation in a hospital setting.</p>","PeriodicalId":76136,"journal":{"name":"Medical progress through technology","volume":"21 4","pages":"177-80"},"PeriodicalIF":0.0000,"publicationDate":"1996-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Medical progress through technology","FirstCategoryId":"1085","ListUrlMain":"","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
The accuracy of the radiology vocabulary domain of the IBM VoiceType Dictation for Windows speech recognition system for use in a clinical setting was evaluated. Six men and one woman dictated a case report under several conditions. The impact of vocabulary domain, microphone position, personal enrollment, and background noise were assessed. The dictations were exported to Microsoft Word 6.0, printed, and graded for accuracy. The data were expressed as a percentage correct. The enrollment and training processes took an average of 5 hours to complete. The accuracy of VoiceType Dictation was significantly reduced by speech volume and background noise level so that it could not be used to dictate clinical information accurately in a hospital setting. Although IBM's VoiceType Dictation with the radiology vocabulary domain has improved speech recognition, it is not a reliable dictation system for accurate and efficient clinical dictation in a hospital setting.
评估了用于临床环境的IBM VoiceType Dictation for Windows语音识别系统的放射学词汇域的准确性。六名男子和一名女子在几种情况下口授了一份病例报告。评估了词汇域、麦克风位置、个人登记和背景噪音的影响。听写导出到microsoftword6.0,打印出来,并对准确性进行分级。数据以正确率表示。注册和培训过程平均需要5个小时才能完成。语音类型听写的准确性受到语音音量和背景噪声水平的显著降低,因此无法在医院环境中准确地口述临床信息。尽管IBM的带有放射学词汇域的VoiceType听写改进了语音识别,但它并不是一个可靠的听写系统,无法在医院环境中进行准确和高效的临床听写。