General discussion


Speech recognition technology: Pros and cons

By debate ·
What's your take on the potential of speech recognition technology in the healthcare arena? What improvements do you think are necessary? How concerned are you about privacy, security, and accuracy? Share your comments about the viability of speech recognition technology, as discussed in the March 30 Healthcare IT e-newsletter.

If you haven't subscribed to our free Healthcare IT e-newsletter, sign up today!

This conversation is currently closed to new comments.

3 total posts (Page 1 of 1)  
| Thread display: Collapse - | Expand +

All Comments

Collapse -

should look at voice verification as well

As the Healthcare problems/solutions continue to grow, the "thinking outside the box" mentality might seem to want to take shape. I have been involved with voice for over 16 years (co-developed the first IVR/VRU for Executone) and quite frankly many of the so-called voice recognition companies back in the early 90's (and even today) have presented some of their technologies/capabilities that did not work, and today many of us have had to suffer (thank God for consulting)due to failed promises, that were not delivered.
From my experiences with consulting with SOME (notice I did not state ALL)of the Healthcare companies they are still in the legacy environment when it comes to technology. If you really want to see a company that has some great technology, readers should check-out an Irish based company (Voicevault) that is the ONLY company in the World who has developed a secure PKI version using what they call Voice Key Infrastructure- or VKI. It is a great application for HIPAA and for doctors who may to send patient files over and receive an encrypted voice back from the other doctor. The costs are EXTREMELY low since they sell this in a hosted environment, and use some of the highest security in the World.

Collapse -

Many Doctors have problems talking

by Endoscopy In reply to should look at voice veri ...

My wife does medical transcription. Many doctors will have no problem. However many doctors have one or both of two problems when dictating. One problem is they pronounce like the write. The mumble and/or slur words together. understanding the word they are speaking is many times very difficult. The other problem is their choice of grammer and sentance structure is sometimes abhorent. A good transcriptionist has to understand this mishmash and create a document that makes sense.

This means that the software to really replace a Medical transcriptionist will have to deal with both of these problems. An account that my wife has been doing is trying to change over to one of these systems. They are having many problems getting the software to just understand them. It does not try to make sense out of horrible grammer and sentence structure in order to make an understandable document.

Collapse -

We use it now

by GSG In reply to Speech recognition techno ...

We use Speech Recognition now. It sounds good at first, but the cost of the system, and the yearly maintenance costs are currently more than what it would have been to have a transcriptionist typing the documents from scratch. The system we have trains itself based on matching what it thinks the voice file is saying to what the transcriptionist types. The system works very well for doctors who speak clearly, dictate all of their punctuation, and don't use a normal.
A normal is a template. For example, an ophthalmologist that performs the same surgery over and over, may dictate something like, use my normal cataract removal with the following, Right Eye, 10cc Versed, no complications.
In that case, it is MUCH faster not to use speech recognition. It's also faster to type the document than use speech if the doc doesn't dictate his punctuation. The vast majority of the docs don't. I think they are on the right track, but until the ROI improves, I don't believe that there will be widespread adoption.

Back to Desktop Forum
3 total posts (Page 1 of 1)  

Related Discussions

Related Forums