Voice Recognition Software in Medical Imaging: The Latest

2012-03-22
 

pacs ris collaboration Say it’s so! That is, say that voice recognition technology is developed enough for doctors, nurses, and medical imaging professionals to use it effectively.

Today’s medical imaging professionals take electronic health records and other digital technologies for granted and with recent advances to technology and capabilities, can high-quality voice recognition be far behind?

It’s mostly a matter of size and processing power. Medical imaging professionals could have the processing power they need to translate their dictation to 99 percent accurate written prose – but it’s more than a bit cumbersome to put the latest IBM supercomputer in a breast pocket.

Programming also isn’t advanced enough to meet the needs of medical imaging professionals. Accuracy rates are consistently above 90 percent now, but that’s simply not good enough. For one, imagine not understanding every tenth word someone says, and then ask whether you want that same problem codified in medical records (or, more problematically, whether you want to pay an army of listeners to make the necessary corrections).

The (literally) millions of nuances in human speech patterns make for slow going in the voice recognition improvement business. But according to an article in Diagnostic Imaging, developers are confident that accuracy rates acceptable to medical imaging professionals will happen sooner rather than later. For one, accuracy continues to go up, even if slowly. Also, processing power continues to go up and size continues to go down, meaning that a voice recognition supercomputer in a breast pocket is a foreseeable reality.

Medical imaging software specialists are also developing voice recognition technology that integrates with EHRs, RIS, and PACS. In other words, they are not just aiming for documentation, but for more medical imaging information immediately available to those directly involved with patient care.

To receive the latest medical imaging news, subscribe to the Medical Imaging Talk blog via RSS feed or email. Or follow us on Twitter.

 

Leave a Reply