At the Society for Imaging Informatics in Medicine (SIIM) 2026 Annual Meeting (June 10-12, 2026), Steve Deaton, Director of Cloud Solutions, Visage Imaging, participated in an #AskIndustry Panel Discussion, moderated by Audrey Verde, MD, PhD, "Under Pressure: Industry Imaging Strategies Rise Above Chaos" along with other industry representatives.
For your convenience, we’ve posted the recorded presentation and we have provided a transcript extract of Dr. Verde’s and Steve’s contributions to the panel discussion:
Transcript Extract
Dr. Audrey Verde [1:50]
So, thank you for everybody being here both in person and anybody who's joining us online. I'm Audrey Verde. I'm going to be moderating this session. This session is called Under Pressure: Industry Imaging Strategies to Rise Above the Chaos. It's an AskIndustry panel. What we're going to do is we're going to go through a conversation of questions about important things going on in our field. Essentially, this session was initiated by the theme that you all are feeling. Radiologists are overwhelmed with work. We're getting burned out. We're looking for ways to optimize efficiency. We're looking how AI can help and we're looking to our vendor partners to how we can work together to solve this problem together.
And so that's what we're going to hopefully have a conversation about today. We're going to chat it out for about 45 minutes and then open up to the audience at 15 minutes left. So you all can help me. I have others helping me, but when it's 15 minutes left, you can just start walking up to the mic to ask questions to give me a signal. We need to cut it off up here, because I want to hear from you too, and what's important to you. Okay? All right. If you want CME or CE for this, you have to first add it to your schedule in the app. Just a heads-up. So go ahead and do that. All right. And we have several great speakers with us today. I'm going to let them go down the line and introduce themselves and then we're going to jump right into questions. Okay? Here we go.
Steve Deaton [3:17]
Good morning. My name is Steve Deaton. I'm the Director of Cloud Solutions at Visage Imaging and I've been in the healthcare industry for a few decades now. Early on I was very product and engineering focused, and today, I get to actually be closer to the physicians and actually the delivery for what makes an impact in their life and their work life balance. And so, that's my day to day, is interacting with some of the nation's best and brightest physician groups, and learning what is important to them. And I'm happy to discuss that today with everybody here.
Dr. Audrey Verde [5:51]
Awesome. Thank you all so much for being here. I hope you all are excited. We have a broad range of expertise here. So we're going to get the party started. So if you all could fix one bottleneck in the radiology workflow, what would it be and how? You first.
Steve Deaton [6:05]
Yeah, sure. I came prepared for some of these questions and getting to participate at SIIM and go to some of the actual presentations kind of pivoted my thoughts. So, yesterday I attended a few of the meetings where folks from Mayo, folks from University of Iowa were presenting. And one common thread in their focus and their messaging was to really define the goal that you're trying to reach. So you bring up a problem and everybody is aware of the blood, sweat, and tears that is the mess of radiology right now, overworked. And a lot of people are talking about how to squeeze more studies into the physician's current crushed lifestyle, frankly. And last year when I was on this panel, Dr. Zaragoza from UCLA was sitting next to me and he said, "When a system works really well, it's like a warm bath." And that goal, I kind of digested that over the last year of, what are we trying to actually do?
And I think we have to define what we want to do and I think that it's mostly about improving the radiologist work-life balance. That's the goal that we need to reach because it's not about can they do more RVUs or work more efficiently. It's can we give them the bandwidth that they're able to actually focus on the patient care more than just the number of studies that they have to read per hour or per day, right? When we see a physician have the ability to have a little bit more free time, they spend more time on cases that need it, and they are able to think more openly about the different diagnosis that are available to them, versus just click, click, click. So I think identifying the goal of what's really achievable to improve the radiologist's work life, is where we need to focus on. And then behind that, there's probably dozens of things that can contribute to that, but identifying the goal I think is most important.
Dr. Audrey Verde [11:35]
I love this. All right. So we have a bottleneck of radiologists. That's the issue. We don't have enough for all the work we're doing, but we need to reduce interruptions. We need to increase the visualization of the data, we need to do our job and we need to decrease the mundane. I love this. Okay. In the world of AI, AI is getting deployed everywhere. How do you all see AI helping with these stressors in the market?
Steve Deaton [12:00]
Yeah, I think workflow focus tools are the most important. So like you mentioned, worklist, I think someone mentioned aggregating the data elements into one spot for the radiologist, before he or she has to open that case and see that something's missing. I think pixel based AI is obviously very up and coming although it's been around in mammography in the form of CAD for well over a decade and obviously has a good purpose, but right now, I think the most traction from AI will be had with facilitating everything ahead of the radiologist opening up that study and even other areas of AI. There's some novel approaches to actually have AI running in the background on every dataset to identify other pathologies that the patient might not even be there for. So AI doesn't just have to focus on improving the efficiency of a radiologist.
I think that there's other upstream applications of AI to identify things in the background and promote other service delivery areas. If you can identify cardiac events because somebody was incidentally in there for another CT, from a car crash or something, using those times to accelerate the patient's benefit is something that I think we all need to make sure we're keeping first on top of mind. Because while we want AI to help the physician, ultimately we want to help the patients the most, right?
Dr. Audrey Verde [19:43]
Yeah. So getting AI, while it has the potential to help, it's really important in how you do it. You brought in essentially data is key. Putting in AI when the foundational data is essentially garbage isn't going to help. Garbage in, garbage out. So what do you all do as vendors to partner with your users to get to a state of standardized data, to then also have a state set up where they're going to be able to monitor the AI and whether or not there is any gains, right? What are the measurements that there's a return on investment? Just interested to all think about that.
Steve Deaton [20:23]
Yeah. I think that's another bit of noise that gets added into the soup. So you've got radiologists that are working in legacy systems often and you've got the IT team that is a bunch of mechanics trying to keep the wheels turning on the legacy system. And then you're now bringing in, let's talk about new AI technologies and new workflow. I think it's important to clean your house first and enable your data for the AI, move to the cloud, get it into areas that it's more accessible from other applications and other stacks. And that's one thing, that the cloud actually lubricates the adoption of AI. And a lot of folks are just looking for the application tier and they're not focusing on the IT underlying layer. And that's one thing I think that everybody needs to prioritize a little bit higher in the path there.
Dr. Audrey Verde [26:27]
Yeah, totally. We already started drawing out cloud components, so that's where we're going to go next. There's lots of people who still have on-prem situations, but we're all like, "Okay, we all need to move to cloud." But there's some hesitation and some concern about cloud. What are you going to do when the internet goes down? Because you've seen that can happen now. What do we do when we're totally hacked? How will you do for performance, speed, cybersecurity, integration? What do you have to say to those concerns? And then, after we talk through the concerns, talk about the benefits of cloud and how it can start helping to address this workforce issue and trying to address efficiency.
Steve Deaton [27:10]
Yeah. I face this every day almost, and the concept of moving your IT infrastructure stack further away from the building that you operate in is intimidating for a lot of people. Your IT teams are mechanics right now, they're keeping the system running and now it's going to be farther away where they can't touch it. But the reality is that the amount of money that these hyperscalers are investing in their data centers, in their connectivity, in their security, is greater than all of us together could invest in our own data centers. So it is the most performant. It is the most connective, redundant, awesome uptime. I mean, the cloud vendors have deployed their own private fiber into every city in the nation. So the concepts of internet outage and that PTSD of variable performance from your internet connectivity actually isn't in the equation, but people assume, "Oh, I'm in the cloud, I'm dependent on these things."
At Visage, we haven't done an on prem deployment in six years and our median customer size is about 1.5 million studies a year. So these are really large organizations doing massive workloads, and the cloud has been a measurable performance improvement and security posture improvement. You don't have to face patching cycles with on-prem. You don't have to worry about those nasty stats that say on-prem devices are 63% more likely to be behind on some security update or patching event. You know, so take advantage of the money that these cloud vendors are putting into their data centers. It will make things better for you. It will make things less hands-on for your IT teams.
And just to pass it off, the AI topic is just further empowered by all of these things. So don't look at the cloud as risk. Look at it as bolting on a supercharger. It's going to just accelerate everything for you.
Dr. Audrey Verde [34:45]
Nice.
Steve Deaton [34:46]
Just to add to that, one thing that we face from public facilities that are going through a rigid RFP process is that there's a very regimented communication cycle and it's limited. And so to Kyle's point, your staff might not have the opportunity to have enough engagement. So one tip I would really recommend that you take your organization is please structure the RFP process where we have more time to have these lunch and learns with [y]our potential vendors. Because if you're not consuming their content enough and you're operating, like Kyle said, with a lot of PTSD concepts in mind, you're not going to get onto the new proper path that you really should be going towards. So that'd be something that we face as a challenge that I would recommend you kind of take internally and improve.
Dr. Audrey Verde [36:00]
They're trying to trip you up. Yeah. Okay. All right. So this next question, we're going to start from the other side. All right, this is SIIM, we are where everybody belongs. We have all the -ologies represented. We have pathology, ophthalmology, cardiology, dermatology, visual light, radiology. But in a space where all the institutions are resource strapped, most places have just a single IT team for the hospital. They're trying to figure out how to use those resources across all the departments, even though I think that's incredibly unfair because radiology is completely informatics and we should have larger proportion of the IT team. But I'm biased. Anywho, but I love all you other ologies out there. I do. I do. I have love for you, but I think we need either our own IT team or a larger bit of IT team. Anywho, how do your organizations address supporting both radiology imaging informatics and enterprise imaging informatics?
Steve Deaton [43:30]
Yeah. I think opening up enterprise imaging has a lot of significant benefits, like you just said, that the data is orthogonal, but it also, everyone needs to recognize it's a major risk point because what you're doing is you're bringing in a lot of other folks who have their own forms of bad experiences and PTSD, right? And if you plot revenue by department, right, cardiology is massive, radiology is massive. If you look at this in a bar graph, it drops off significantly in other ologies when you look at their imaging contribution, right? And I think that from a C-suite level, they need to look at imaging from that lens as well. Because it doesn't necessarily make sense to have equal voting members on a decision committee from a team that does almost no imaging and is, for lack of a better word, kind of riding coattails on the great informatics work that the radiology department is accelerating for the organization.
So that is a distraction that I see occur with different partners of ours and I'd encourage you to get ahead of that and manage into that, because ultimately for the patient and for AI, all the data needs to be in one system. That's the proper path. But from a decision making perspective, too many cooks in the kitchen, so to speak, I think we all need to remember that radiology is the significant contributor, materially. We have one prospect who's facing closing over a dozen of their rural hospitals for financial implications in the upcoming years and actually increasing their spend in radiology software will increase downstream revenue. And that alone is going to allow them to subsidize and keep these other dozen plus hospitals open.
So it's really a mind shift to say, we're going to overinvest in radiology because of all the other benefits it's going to have downstream for the physicians, for the patients, for the communities and whatnot. And so the enterprise imaging topic is a really sensitive one and it needs to be managed with care and with focus. I think you really have to focus. Too many of the IT groups, they're just as overworked as radiologists, right? And they're spinning so many plates and what the mind shift that we're constantly, I think all of us are trying to push into our partners is, when you move things to the cloud, when you implement standards and adopt these interoperability standards, things get easier. You don't need the mechanics and you can go on the offensive now and say, "How do we improve the workflow versus just keep the doors open and keep the engine running?"
Dr. Audrey Verde [46:46]
So what I'm hearing is the concept of enterprise imaging is important. We need to get all the imaging into one location, try to harmonize the vendors across, or to collaborate, or be interoperable across an enterprise. The problem is getting the resources to do that and how to structure those resources and honestly how to convince the C-suite to do that and to understand that most of the times it's not radiologists up in the C suite. Most of them are not even doctors, let's be fair, right? They're MBAs, right? So how do we translate that knowledge? How do we translate this need so that we can work forward as across all fields, across all the ologies working together to lift up radiology will actually lift up the entire enterprise and make enterprise imaging more possible.
Dr. Audrey Verde [48:59]
I love this. You're just teeing it up for me. I'm about to go to standard space. All right, so yeah, coming from RadLex and LOINC, and as a IHE radiology committee member, I'm interested in hearing how your organizations are involved in the standards' creation, standards implementation. How do you let the community know which standards you are using, which IHE profiles you support? Are you engaged in updating the IHE profiles? If not, can we get a conversation as to why? And yeah, essentially we're trying to work together. We're all trying to solve the same problem. The IHE profiles and standards exist for interoperability to work. So let's have a conversation about how we can all do that together and what you're already doing.
Steve Deaton [54:00]
I have to comment about it. I totally agree with what you guys have said and what everyone needs to really understand is that probably a good number of the vendors that you'll be considering, they already support all these standards. Visage does the same thing as like what Kyle was saying. We go to the different Connectathons and we manage all the profiles and make sure that we're meeting the newest standards and whatnot, but the problem is actually adoption by the customer, adoption by folks such as y'all and your organization. So we've supported FHIRCast for many years in high volume environments, but we have very few customers that are actually ready to go and deploy that technology. My observation is that it's been because they're spinning other plates, keeping the doors open with other problems, and that's why I've focused my career at this point into cloud, because although I've been in healthcare for 20 years, cloud is the current enabler for adoption and deployment of the standards that already exist, that we're all staying ahead of the game on making sure they're there.
We need demand from you all and that means you need bandwidth to be able to focus on how you could make a better experience for the physicians or the patients, and that's where we need y'all to spend a little bit more time and focus because the interoperability is sitting there on the shelf ready for you. We just need you to push those workflows forward.
Dr. Audrey Verde [55:55]
All right. So we have a question. Oh, mic's not on.
Audience Question [55:38]
Okay, so thinking about interoperability and standards, right? We have a vendor neutral archive. Awesome. It can take non-DICOM studies or data images, take a PDF, import it. Great. Is it enough to say, "Hey, we can ingest non-DICOM data now and store it in the VNA," or should we really be looking at those workflows and instead of saying, "Yeah, great. It's in the VNA now." Should we be going back to the vendors and saying, "You really need to be DICOM or some other standard format rather than just, "Hey, I ingested it and it's in my universal viewer. Cool."
Steve Deaton [56:21]
I'd say why? What's your goal? Got to define the goal before you can get to the right answer and I'm used to seeing this kind of discussion and I think it's what outcome are you chasing with a priority and then that kind of dictates the conversation from there on out. It's not about focusing on if we can do it, therefore let's do it. It needs to be take it on a little faith that it's probably possible in one way or another. Why? What's your goal that you're trying to target? Because if you don't get that and you don't clearly target that, your physician base, your champions, your C-suite, all those people aren't going to be able to align and then it's going to have struggle to adopt throughout the facility.
