Expressing appreciation for the technological advances that have limited previous disruptions to radiology workflow, this author speculates about what could come next with AI, voice recognition and workstation enhancements.
Arthur C. Clarke famously wrote “Any sufficiently advanced technology is indistinguishable from magic.” A lot of folks have examined and expanded on the concept. I’m surely not going to say anything new about it here, although I will refer to its subjectiveness:
Show people 2000 years ago a working light bulb, for instance, and they might consider you a wizard. Today, you would have to pull a much bigger rabbit out of your hat — teleportation, maybe, or fabricating something out of nothing.
The trajectory of advancement is usually a one-way street, barring societal upheaval and intervals such as the European “Dark Ages.” Occasionally, however, we get a little backwards movement, after which the restoration of what we had previously taken for granted can feel borderline magical.
I have had more occasion to reflect on this than the average radiological bear, courtesy of a recent blog (see “Keys to Defusing Misperceptions About a Radiology CV of Great Lineage”). Different gigs come with varying levels of technological sophistication and my current rig is literally a decade ahead of some other systems I have worked with in my career.
That might not seem surprising, since I finished residency in 2004, but I will add that a relatively recent employer ran Windows 7 on its devices. (Google tells me that OS became available in 2009 and was wound down in 2015. The job in question was years later.) You may imagine that an outdated operating-system wouldn’t run well with more current software and hardware, and you would be right.
Keeping such suboptimal tech in mind, it is pretty easy to feel a strong sense of gratitude at all of the comparatively magical ways my current workstation operates. Here are a few off the top of my head …
* My workstation boots up practically instantly rather than humming and whirring for several minutes.
* It almost never requires a reboot due to crashing software (or because unmotivated IT support people offer no other suggestions), but when it does, it is also instantaneous. Rather than “Do I really have to reboot? That’ll eat a chunk out of my day,” it’s a non-issue.
* It doesn’t have multistage log-in procedures that (surprise!) crash more than occasionally. (A simpler log in is also a perk of working from home, where nobody has access to my machine but me.)
* My workstation can run more than a couple of programs at once without bogging itself down. I don’t have to think about whether opening another browser window is worth the risk of derailing everything else.
* Unlike some other rigs I have seen that were “virtual machines,” having to transmit all transactions between states via Internet (and slowed down by VPN, another encumbrance my current system is without), everything is actually in the room with me. Even with high-speed connections, the absence of that lag feels like I have moved from slow motion to being the Flash. Plus, I am pretty sure it was playing havoc with my voice recognition accuracy.
I am optimistic and realistic enough to see that there is more magic yet to come. Much as I appreciate my current state of affairs, technology marches on, and I have still (hopefully) got a couple decades of work left in me. I can sit here today and try to dream up the improvements of tomorrow, but the real sense of magic comes when some new innovation comes out of left field, wonderful stuff for which I wouldn’t have dared hope or even conceived.
On the more pedestrian side of things, hope springs eternal that voice recognition will become the reliable tool that it was promised to be, something I can count on to get the job done all of the time, without constantly having one of my eyes nervously on the dictation screen to catch innumerable errors.
Maybe artificial intelligence (AI) will be up to the task, and actually learn my customary words, phrases, and formatting as opposed to the flimsy promises of voice rec “learning as you go” that I have been hearing for 20 years. Customizable voice commands for routine actions would be nice too, a la “get the referrer on the phone.” While we’re at it, let AI watch me operate my RIS/PACS, and offer adjustments to the software and its settings to better match my habits.
I would, for instance, appreciate the software noticing that I mentioned a lung nodule, and quietly opening a small window showcasing the Fleischner guidelines and Lung-RADS in response, preferably somewhere that didn’t superimpose my images or dictation area if you please. I have perhaps a dozen things that I constantly look up just to make sure I remember their details correctly. They would be just as welcome to materialize when appropriate.
I also dwell in enduring hope that, someday, I will return to the most efficient setup I ever enjoyed for input at the workstation. This would include a left-hand counterpart to the mouse my right hand is wielding. Keyboards and gaming-style keypads only go so far.
A shade more magical would be something I wrote about a long time ago in this blog. How about a virtual reality workstation? (Holograms and input gloves could work, too, shades of Minority Report, but I think that would be more expensive). It could get rid of the MonitorHenge that clutters up a typical rad’s desk, and for that matter let him or her do the work in a living room recliner just as easily as an old-school office.