SIIM: Scoring a slam dunk with speech recognition

MINNEAPOLIS—In the last decade, speech recognition has blossomed from a bleeding edge to leading edge technology. “Today, the question is not do I need speech recognition but how do I make best use of it,” stated David Hirschorn, MD, director of radiology informatics at Staten Island University Hospital in New York City. During his presentation at the Society of Imaging Informatics in Medicine (SIIM) conference last week, Sirschorn said speech recognition was a ‘slam dunk’ at the hospital.

The deployment slashed annual reporting costs from $300,000 to $50,000 and cut report turnaround time from up to two weeks to less than 10 minutes. Speech recognition has the shortest return on investment of any major imaging informatics projects, said Hirschorn. Within six months of deployment, the department eliminated its seven full-time transcriptionists, and 85 percent of reports were completed in the speech recognition system.

Despite its impressive advantages over transcription, speech recognition is a "tough transition," said Hirschorn, who provided a brief overview of best practices in speech implementation. These are:
  • Identify a physician champion. Speech recognition is not as easy a sale or deployment as PACS. Having a champion ups the odds of acceptance and ownership among radiologists.
  • Explain and re-explain the benefits of the technology. “If radiologists are not miserable in the first stages, something is wrong,” said Hirschorn.
  • Invest in training. When radiologists learn how to use the system well, they can be more efficient in the long run.
  • Utilize macros and autotexts wisely. Speech recognition makes radiologists slower in the beginning. Macros and autotexts ultimately accelerate speed and provide a foundation for structured reporting.
  • Optimize for low bandwidth as needed. Physicians need to use the system from multiple locations including home. Making sure speech works well in low bandwidth situations helps boost acceptance.

There are multiple challenges that new speech sites may encounter, said Hirschorn. For example, some physicians may be prone to over-editing, which wastes time and decreases efficiency.

Training and support should minimize this tendency, he said. Despite improvements in the technology, foreign accents are tougher to manage. There may be a few physicians who cannot use the system because it cannot understand their language usage. Finally, speech, like PACS, becomes a single point of failure for whole radiology department. Developing a business continuity plan is a critical part of the project.

“Speech recognition will continue to grow because it makes sense clinically and financially. It is a difficult, but worthwhile, transition,” summarized Hirschorn.

Around the web

The American College of Cardiology has shared its perspective on new CMS payment policies, highlighting revenue concerns while providing key details for cardiologists and other cardiology professionals. 

As debate simmers over how best to regulate AI, experts continue to offer guidance on where to start, how to proceed and what to emphasize. A new resource models its recommendations on what its authors call the “SETO Loop.”

FDA Commissioner Robert Califf, MD, said the clinical community needs to combat health misinformation at a grassroots level. He warned that patients are immersed in a "sea of misinformation without a compass."

Trimed Popup
Trimed Popup