SR Vendors Talk

Whant ews peecrhe cognititoenc hnologiesa rep hysiciarnesq uesting? Finke: Physicians are less inclined to work around technology shortcomings and prove the benefits of "typing with their tongue" at their expense. The focus is on behindthe'scenes use of more comprehensive technologies (no training or workflow disruption required) that allow them to be more productive while delivering higher value reports (e.g., text and clinical data). Their requests are an emphasis on integration into existing workflows and allowing for flexibiliry of hybrid review processesin volving either medical transcription and/or physician self-edit. Fleming: Physicians want technologies that eliminate delays and inaccuracies in communication to improve patient safety and streamline reporting. Radiologists want to be as productive as possible and are requesting solutions that deliver all the needed tools through one workspace. Among other things, the demand is for systems that create reports with voice-driven "shortcuts" and automatic "normals," which send critical messagesto ordering clinicians instantly and offer diagnostic decision supporr via simple voice command. Radiologists are looking for one solution that delivers speed and accuracy and allows a seamlessin tegration within their worksoace. Whaat res omeo bstacleasn ds olutions regardinsgp eecrhe c0gnititorna ining and productivity? Finke: Conventional front-end speech recognition requires significant changes in speaker behavior and workflow to "adapt rhe physician to the technology needs." The side efFect:e xtensiveu ser training required and reduced productiviry. The outcome: Iow adoption rate. "Speech understanding," on rhe other hand, applies a unique combination of speech recognition and natural language understanding to transform dictation automatically into accurate) completely formatted, encoded clinical documents that require no training or change of workflow, and physicians dont need to know about it. Fleming: Scheduling training with radiologists is a challenge. Providing one-on-one training maximizes their time and teaches them how to use speech recognition within their environment. By providing radiologists with customization and system shortcuts, they are empowered to optimize their workflow as they see fit and do things like use speech recognition in batch or real-time mode; control and customize how report information is automatically formatted; and edit speech-recognized reports with voice command, microphone controls or a keyboard. The goal is not to change their workflow unless it's going to make them more productive. ffi Michael Finke is CEO of M"Modal, Pittsburgh. Robert Fleming is senior product manager of Dictaphone Heahhcare Solutions, Burlington, Mas.