I had the opportunity to observe Dragon’s speech to text software utilized in a clinical setting. For those of you who are unfamiliar with this software, it basically allows doctors to dictate their notes by talking into a microphone. The software then converts the voice into text that it pastes into whatever text editing screen you are using. To be honest, I was really impressed with it. I’m not sure whether the software learns medical jargon or whether it’s optimized for the clinical setting (my guess is the latter). Either way, it does a wonderful job of understanding dictated medical speech.
For instance, it immediately knew to write PSA, A1C, TSH, and other medical jargon I wouldn’t expect it to know outright. It’s clearly a time saver; instead of wasting time typing and hoping that you don’t make a mistake, the software is efficient and accurate. However, I did notice that there seemed to be a delay between dictation and text writing. I wasn’t sure if that was designed on purpose to wait for the entire statement before proceeding or whether it was due to stress on the computer system.
Additionally, I was curious as to why doctors need to first see the patient and then catalog the meeting. Why can’t each room be fitted with a microphone, so that the software writes the entire encounter into the patient’s file? That way a record of the clinical encounter is saved the way it truly occurred. I imagine this would be useful in a legal sense as well—having a record of the actual transcript rather than a doctor’s recollection would be tremendously useful. This would save the doctor time as well because he wouldn’t have to go back and spend time recounting the clinical encounter.