TMCnet - World's Largest Communications and Technology Community



IVR - Medical Imaging with Voice Recognition Software Develops
November 04, 2011

Medical Imaging with Voice Recognition Software Develops

By Rachel Ramsey, TMCnet Contributor

According to a recent poll by Diagnostic Imaging, users of voice recognition software find it error prone and inconvenient. Despite the software’s record of reducing report turnaround time and its promise for populating and mining structured reports, 30 percent of the respondents who use the software reported frustration with it.

Radiologists have high demands for the software because of their need for accuracy, fast-paced talking and use of complex terms. The mathematics for the software hasn’t changed much over the years but the accuracy has improved by over 15 percent, from 75 to high 90s, over 10 to 15 years.

Some radiologists need to be convinced of the software’s benefits, as it needs to fit into their workflow seamlessly and provide data mining tools to make the effort worthwhile.

“We still do encounter some areas of resistance or continued inquiry or challenge,” said Don Fallati, senior vice president of product management for M*Modal. They are “proving or demonstrating that it is realistic and productive that it meets the usability test.”

However, most radiologists embrace the voice recognition because it has the ability to make their lives easier. As the software continues to evolve, opportunities will grow in how the amount of data dictated is mined and shared. Said Petro of the future of voice recognition, “it’s beyond speech.”

As the software continues to evolve, future systems hope to be automatically populated in templates with patient data, saving the radiologist time and effort in the dictation, according to Jacques Gilbert, global marketing manager for radiology for GE Healthcare.

“It’s a bit of a frustration for the radiologist to have to do that when they can see the data is in the system,” he said. “There’s a tremendous amount of time savings when the data they’d like to see in the report from the RIS or PACs are in the report in a structure that they have decided on for different procedure types.”

Nuance (News - Alert) Healthcare has been improving the error rate by fine tuning the speech recognition algorithms and taking advantage of the computer chip power.

“Radiologists are trying to provide an enormous amount of documentation in a very short period of time, accurately, and they are experts at grinding through that workflow and process as fast as they can,” said Joe Petro, senior vice president of research and development for Nuance Healthcare. “They hold us to a much, much higher standard.”

The technology has gotten better at differentiating natural pauses in dictation and accents, said Gilbert. GE Healthcare’s algorithm takes into account corrections a user makes, and the GE tool learns from its mistakes and applies changes made. The time it takes to make the fixes is the main feature where products compete, Gilbert said.

Competition for the products is also based on the product’s account training time and accuracy as soon as it’s opened, said Time Kearns, GE Healthcare’s product manager.

Some of the frustrations by early users might be attributed to the training time. If the software is making the same mistake repeatedly, there are features that allow the user to modify it to make the mistake in the future. Users who find the software frustrating may not have spent the time to understand the frustrations.

Vendors are exploring how to develop systems that can really understand what the radiologist is saying, and populate health IT systems accordingly through natural language processing, which takes the unorganized narrative, codes and structures it, and harvests it for specific data.

The goal of natural language processing is that items, such as patient’s problem list, can be automatically sent to the electronic health record, or certain decisions can be automatically checked against appropriateness criteria.

In related news, according to Stanford University Professor Clifford Nass, female voices are better for IVR systems. It’s easier to find a female voice that everyone like than a male voice that everyone likes.

Rachel Ramsey is a TMCnet editorial assistant, contributing news items and feature articles on a variety of communications and technology topics. Rachel has previously worked in PR and communications at The Wriglesworth Consultancy, an award-winning London PR firm. She has also contributed to the creative services department at CBS 3 and The CW Philly in Philadelphia. To read more of Rachel's articles, please visit her columnist page.

Edited by Juliana Kenny

More IVR Feature Articles

Technology Marketing Corporation

35 Nutmeg Drive Suite 340, Trumbull, Connecticut 06611 USA
Ph: 800-243-6002, 203-852-6800
Fx: 203-866-3326

General comments:
Comments about this site:


© 2018 Technology Marketing Corporation. All rights reserved | Privacy Policy