23 September 2011
Breast imaging reports generated using an automatic speech recognition (ASR) system are six times more likely to contain errors than those created with conventional dictation transcription, a new study in Canada shows.
Lead researcher Anabel Scaranelo MD, of the University Health Network in Toronto, found that 23 per cent of automatic speech recognition had more major errors compared to four per cent of conventional dictation transcription reports.
The study focused on 615 reports of complex cases which had been discussed at disciplinary level.
Dr Scaranelo said that the error rate was even higher if breast MRI reports were looked at separately and also that native language had no effect on results of the voice recognition device.
She added: "We thought that there may be a higher error rate for non-native English speakers because the software works with voice recognition, but that didn't happen."
In 2008, 48,034 people in the UK were diagnosed with breast cancer, according to Cancer Research UK.
By Edward Bartel
Scaranelo, A. et al., "Original Research: Error Rates in Breast Imaging Reports: Comparison of Automatic Speech Recognition and Dictation Transcription", American Journal of Roentgenology, October 2011.
Health News is provided by Adfero in collaboration with Spire Healthcare. Please note that all copy above is ©Adfero Ltd. and does not reflect views or opinions of Spire Healthcare unless explicitly stated. Additional comments on the page from individual Spire consultants do not necessarily reflect the views or opinions of other consultants or Spire Healthcare.