Related Experiment Videos
Voice recognition for radiology reporting: is it good enough?
D S Rana1, G Hurst, L Shepstone
1Department of Radiology, Norfolk and Norwich University Hospital, Norwich, UK. sameen.rana@nnuh.nhs.co.uk
Clinical Radiology
|October 15, 2005
Summary
Voice recognition (VR) offers faster radiology report production than traditional dictation-transcription (DT). However, VR shows a tendency for more errors, particularly among less experienced users, though it is viable for experienced professionals.
Area of Science:
- Radiology
- Medical Informatics
- Health Services Research
Background:
- Traditional tape dictation-transcription (DT) is a standard method for generating radiology reports.
- Voice recognition (VR) technology offers a potential alternative for improving reporting efficiency.
Purpose of the Study:
- To compare the efficiency and accuracy of radiology reports generated by voice recognition (VR) against the traditional tape dictation-transcription (DT) method.
Main Methods:
- Two hundred and twenty computed radiography (CR) and cross-sectional imaging (CSI) examinations were reported using both VR and DT.
- Reporting times and errors were compared using univariate analyses and a general linear model.
Main Results:
- Voice recognition significantly reduced overall report production time for both CR and CSI.
- However, VR use was associated with a significant increase in total and major errors for CR, and total errors for CSI.
- Less experienced users showed significantly more errors with VR, while experienced users did not.
Conclusions:
- Voice recognition is a viable reporting method for experienced radiologists, offering quicker production times.
- Inexperienced users may experience a higher error rate with VR, necessitating careful consideration of user experience.