StructuRad Speech Recognition Architecture
What Works and Why Dual-speech Engine with Dragon NaturallySpeaking is Better
StructuRad uses an innovative dual-speech engine implementation to achieve the best-of-breed performance for both voice-enabled command-and-control and continuous speech recognition.
This unprecedented level of accuracy will redefine how the market measures usability. Have you seen a sales person handing their microphone over to a prospect and the system just works?
Yes, our software does that, even without changing user profiles! Why is the StructuRad implementation better? Rather than just relying on the traditional speech recognition algorithms,
domain-specific lexicon and user profiles.
Dual-engine Speech Recognition Radiology Reporting
This architecture leverages the additional power of a comprehensive knowledge base, massive macro library, computation linguistics, GuidepostTM templates and Speech Studio's
speech operating system to manage two speech engines. This approach allows our voice command-and-control function to continuously refine the dynamic grammars in narrowing the context for
superb recognition with the benefit of a recognition-mechanism that is speaker independent, i.e. it works without any user profile. Hence, it alleviates the need for time consuming training.
Continuous Speech Recognition Dragon NaturallySpeaking
We integrate Dragon NaturallySpeaking profile-driven speech recognition with a medical lexicon. The architecture of using the Speech Studio speech operating system can unify the VUI (voice user interface) between us
and other applications. In other words, our speech recognition mechanism can easily extend to help you with voice-enabling your PACS and RIS.
© Copyright 2002-2008 StructuRad LLC, All Rights
Reserved.
StructuRad, FastMacros, Macro Dashboard, PACSPal, StudyManager, RadBoost,
and ReportNow are trademarks of StructuRad LLC. |
Privacy and
Legal Information
Webmaster E-mail contact |