This transcription is provided by artificial intelligence. We believe in technology but understand that even the smartest robots can sometimes get speech recognition wrong.

Newsday: Imaging AI Expansion and Apple-Only Hospitals with Chris Jenkins

[00:00:00] This episode is brought to you by Healthlink Advisors. Value. Insights and Solutions. Expert consultants serving the healthcare industry. Check them out at thisweekhealth. com slash Healthlinkadvisors.

Bill Russell: Today on Newsday.

Chris Jenkins: We've always talked about look backs in healthcare when you go buy something, but I think this is critically important to maintain that

look back in any type of AI solution you're putting in place regardless of the domain you're looking at.

Bill Russell: My name is Bill Russell. I'm a former CIO for a 16 hospital system and creator of This Week Health. where we are dedicated to transforming healthcare, one connection at a time. Newstay discusses the breaking news in healthcare with industry experts

Now, let's jump right in.

(Main) Alright, it's Newsday, and today we're joined by Chris Jenkins with Healthlink Advisors, EVP of Enterprise Imaging and Digital Strategy. Chris, welcome to the show.

Chris Jenkins: Thanks. Looking forward to the discussion.

Bill Russell: A lot of things going on. [00:01:00] let's start with security. So kettering Health had a system-wide technology outage last week. I think it might still be going on if I'm not mistaken. I've not checked this morning, but I know that last week, the 23rd, 24th, 25th, I'm looking at their page right now. It looks like they were almost completely offline. The interesting thing about this, and I'm curious, your take on this, is it's not surprising to me that people get in, it's surprising to me that they're still able to move laterally and shut down an entire system.

Chris Jenkins: That remains a surprise to me as well, Bill. And I think, the way I look at it is, once they're in, how do you contain them to some extent, but what's the overall impact to the business? And it comes down to the business resiliency. , How do you still function?

How do you still deliver care? How do you still manage the financial side of your business while IT is trying to recover? This one's interesting though because in addition to this, there were scam messages that were going out to [00:02:00] patients. So this is a new added addition to these type of outages.

So, the infiltrators must have gotten a hold of some type of information to go out and access patients because all communications was down. So, think of that your mom, your grandma needing an appointment or having a surgery and they show up.

They can't have the surgery and there's no communication out to let them know what's going on. So, it's kind of the intersection of business resiliency. How do you maintain operations? Who owns that piece of it as well as, how do you recover from that and then prevent that from happening in the future?

Bill Russell: 2024 to me was the year. Of talking about security, like we've talked about it incessantly. Yeah. I mean, we had, Change healthcare. We had the CrowdStrike itself causing, Yeah, an issue. So we talked about cybersecurity for most of 24. It feels to me like we just turned the page and start talking about AI.

Like every conversation starts with AI, ends with AI, and I'm like, yeah, wait, where are we on? I throw it into our, our 229 city tour dinner. I'm like wait, where [00:03:00] are we at on cybersecurity? Have we identified all those single points failure or the third party risk, like change healthcare? I'm like, Hey, what was the result of that?

Where have we gone with that? And, yeah. it feels like we've moved on from that conversation a little too quickly.

Chris Jenkins: Yeah. I would agree because it's no longer one system downtime. Right. The EHR used to go down. Okay, that's fine. We have downtime procedures, but I. How do you watch the House 24 by seven?

How do you do everything you do with healthcare organizations that, it's a huge investment, right? To manage those environments and prevent single points of failure, and even if there's a single point of failure, how do you prevent it from propagating throughout the organization where you lose the entire, your entire technical infrastructure?

It has to be a daily thought, in my opinion, because. There are some bad characters out there that all they want to do is figure out how to access your systems and disrupt, the health systems and you can't forget about it. It needs to be top of mind.

Bill Russell: The blast radius still is an [00:04:00] interesting conversation to me.

I did talk to a cybersecurity professional last week and I was asking him about that longitudinal, being able to go Laterally across the network and he said, there's a lot of things that are designed to go across the entire network. I said, well, gimme an example.

He said, active directory. I. Yeah. I'm like, oh, yeah, that makes sense. He goes, well, they'll come in and ride the active directory train across the entire network. We still have endpoint security challenges as well across the board. And I'm like, so what are we doing about that?

He goes, well, we're starting to segment our active directory to the point where you can't ride it across the entire. End up at these choke points where we can identify malicious activities and shut 'em down within a certain sphere. And I thought that was interesting as well.

If nothing else, we'll keep bringing up the conversation around cybersecurity. I'm sure it's top of mind. The people I talked to last week, it's top of mind for them. It's just that the AI conversation is so [00:05:00] loud and why do you think that is? We're gonna look at a story.

We'll, we'll look at the, So Rad AI announced strategic investments from four health systems. Let's see, Advocate, Memorial Health, Corewell, Atlantic Health and this goes right into your enterprise imaging role. Why is the AI conversation so loud? And what does it mean for enterprise imaging that four key systems like this make this kind of investment?

Chris Jenkins: That's a great question. It came out as a shiny object, right? So, when radiologists go to RSNA, there's a lot of shiny objects there. That's how it started. It's still. A couple years back, two and a half, three years when it started a shiny object. But it's actually, it's a reality now.

I think. There's operational efficiencies there are clinical efficiencies of leveraging ai, especially when it's attached to imaging environments. And I think that's why I think it's gone past the hype curve, and , it's where organizations really see it as a little bit of a game changer and it integrates in with their systems and it makes the radiologist much more efficient.

And that's important especially [00:06:00] with a significant, shortage in radiologists, the attractiveness of a modern enterprise imaging environment along with AI tools on top of it to make them more efficient. That's attractive to keeping radiologists. In your health system or at least contract with your health system.

And it makes them better at their job. Now it doesn't replace what they do, there's still the human factor you have to have, but I think that's why there's such a shift around focus on ai. The scary part is if you don't put some type of structure and governance on top of it to understand, what's real, what does work versus what's the value of the solution you're looking at and how do you vet it?

How do you look at bias that's built into it? How do you vet it after post go live to make sure it's still doing what it needs to do? And don't rely solely on that as your solution. So, I think that's a big driver bill. It's just the fact that, there's some real gains in leveraging AI solutions, especially in imaging.

Now, in other areas there are too, but I think with imaging specifically, there's some [00:07:00] proven results in that.

Bill Russell: I'm just looking at the Rad AI website. Just get more familiar with it. Rad AI impressions, they're saying saves radiologists 60 minutes per shift.

And reduces burnout as a result? Well, you would think if it's reducing it by 60, 60 minutes per shift,

Chris Jenkins: Probably an eight to nine hour shift. So it's an hour. That's a big game. Right? And you just think of impressions. So what does that do? So, radiologists, when they go through, they do their reads, they actually document their findings and then the impressions kind of summarizes it.

So at the bottom, here's exactly The finding and the impressions of that finding. So it basically just creates, those impressions for them. They manage it, it creates it in their own language the way they speak, the way they write. And then they just take a look at it and validate it and it's not just Rad AI. There are other AI solutions, but obviously the investment in Rad AI. Another on the clinical side is incidental findings. I kind of say, all right, if you're not looking for Waldo, you're looking for something other than Waldo, but you find Waldo, right?

So [00:08:00] the solution identifies maybe you're looking at a chest x-ray and it finds a nodule or something like that versus pneumonia. It's just something when you're not looking for it and then it's there. It finds it and points you to it. And there again, those are pretty significant functions that help one drive better identification of, maybe a patient event, but also allows the radiologist to be more efficient.

Bill Russell: What does it mean when an, Advocate, Corwell, Memorial Hermann, Atlantic Health System makes that investment? I mean, I would assume, I mean, if I'm the CEO of Rad AI, I'm sitting there going. Hey, validation. Not only have they purchased the product and believe in the product, but they believe in it to the point of, Hey, we want to be an investor.

Chris Jenkins: It's not just an investment. It's a partnership, and they understand the value of it. I think. I kind of wanna say in the beginning, it's attractive, right? So now they have a tool and function that's been proven. It still needs work. I mean, there's still needs some development, but I think they solely believe that it's going to enhance how they do business.

It's gonna [00:09:00] enhance their patient care. So they view it as a partnership now, more so as just an investment but they're gonna see our eyes on it

Bill Russell: so AI is really fascinating to me. It's dramatically changing things. Google I/O last week put out a whole bunch of different tools and I'm looking at those tools and it's, there's so many of 'em.

It's hard for me to get my arms around. Like, I see the video one and I go, okay, that changes this game. And then I just look at the various tools or, and I'm like, alright, that changes advertising. That changes. One of 'em I was looking at going, I think that's gonna change the entire checkout process for online ordering, and I think that's what they want to do.

Yeah. And it's AI driven, so it's just like, AI driven with agent to agent and MCP in the backend. So it's these AI agents talking to AI agents that are operating against a set of doing some functions and just returning back to a LLM type interface of, Hey, your order's been purchased, it's on its way, it's been charged to this credit card, and thank you for your order.

And you're like, man, I didn't [00:10:00] even go to the page and type in the 50 things that they want me to type in.

Chris Jenkins: It's scary. It's there's, there's a, it is funny, there's a book data in Goliath, which is kind of the basis of gathering data and profiles. And everyone, if you think of AI a is really just data driven, now the scary part of that is at times, how accurate is that data? But, it gets smarter and smarter. It's just automating a lot of those manual functions. But data is at the core of a lot of what's going on with AI and building on top of that it kind of propagates even in our business, when we do meetings now we use Gemini to summarize the meeting minutes.

And that's a very simple productivity tool for us. But instead of someone typing minutes or creating agendas, and it can even create agendas for you. There's tools called presentation AI. It will create PowerPoint presentations for you based on a couple simple statements you put in there.

I need a PowerPoint presentation for A, B, and C and it rolls it out. It's it's pretty interesting, but I kind of go back to what I said in the beginning. You still have to put some thought around anything you're doing with AI, [00:11:00] whether it's on the clinical side, the productivity side, ERP.

What are you getting? Is it really accurate? Is it adding value? And it isn't one and done you have to vet that going into it, vet it while you're live on it and post go live. So you just have to make sure you're maintaining it more so than, we've always talked about look backs in healthcare when you go buy something, but I think this is critically important to maintain that

look back in any type of AI solution you're putting in place regardless of the domain you're looking at.

Bill Russell: It'll be interesting to see if MCP servers change how the EHRs function, because in theory you could really change the model. You could have the EHR act as an MCP server, in which case it can receive requests from AI models. Look at the actual data that's in the EHR and then provide a response. Yeah. And people are worried about, oh, you're gonna expose all the data. Well, that's clearly not how it works. It's still working against APIs. It still requires security.

It can limit what data it has access to or doesn't have access to. [00:12:00] But if you think about that if they created an MCP server that, I don't want to use the word exposed after we talked about security but for a programmer, that's the word, right? It exposes Yeah.

That data to an MCP server or LLMs that changes the nature of how we interact. With technology in the hospital. I could very well see a fundamental change in the user interface for how we get at data and potentially even how we input data. I mean, we're seeing, ambient intelligence is probably the fastest growing.

Adoption I've seen of, of any technology.

Chris Jenkins: It is, and it is funny you bring up ambient technology. So I was at a meeting and I was talking to a CIO and he was actually talking about, this is ambulatory practices, they drive a lot of business through referrals, to other specialties.

And the concept there, which I think is a simple concept, is if I have ambient listing in a room when a doctor is meeting with a patient, listening for a trigger that says you need a referral. And [00:13:00] if I can actually trigger that referral real time while I'm there. And in addition to that, have a room set up in the same building, connecting directly to the specialist.

I can actually walk them down the hallway, put 'em in the room, and have them connect with a specialist. That referral at the same time all through ambient listening to trigger stuff like that. But it's pretty interesting, the value there. Now there's probably some privacy stuff you have to think through but conceptually, it's pretty interesting how it could work.

Bill Russell: And that's the point I get back to on this AI stuff. Conceptually you play with something, you go and oh my gosh, that has like 55 different ramifications within healthcare.

Chris Jenkins: Yeah.

Bill Russell: But you have to conceptualize all these things. And what I'm finding when I'm in the room with these CIOs, CTOs and CISOs, is.

It's identifying the use cases that they're going to adopt, and I think the primary driver of the use cases they're going to adopt are what platforms are they on today? They're just waiting for their platform players to say, this is what Epic is doing, this is what ServiceNow is doing

[00:14:00] this is what Workday's doing. Yeah. This is what Microsoft's doing, and they're just waiting for those tools to come out. And to be honest with you, I'm a little disappointed. Because it's like they've been given all these tools and they're sitting there going, well, they're gonna tell me how to use the tools.

They're gonna tell me how to use this. And I'm like I get it. I understand we're really busy, but I think there's a lot of opportunity we're missing just by not I don't know, engaging in just, just dialogue with different groups, like you just mentioned, referrals.

The problem with referrals isn't initiating the referral. The problem with the referrals is when the referral gets over there, it literally goes into this queue and someone reads it, and then in some cases they read it and then they type it into another system and I'm like, oh my gosh. Yeah, that is an AI dream We could transform that overnight if someone would take the time to sit down and look at that workflow.

Chris Jenkins: Yeah. It kind of goes way back to the old macros, right? When you just automate a manual task, right? And prevent that from happening, aI takes it to whole different level. And you kind of [00:15:00] mentioned the beginning Bill, that, I'll say conceptually again, the fact if you have all these different APIs and stuff sitting on top of your EHR, you're.

Imaging platforms, all those other platforms, and you figure out how to pull all that information together. But how do you do that? So I think there's still a traditional mindset, and part of it is keeping your environment secure, but also manage your environments somewhat. You said, okay, epic first, enterprise imaging first your ERP first mindset, which isn't a bad mindset, right?

So you wanna keep your environment as simple as possible with healthcare, so you can manage it and deliver services from an IT perspective, but at the same time, how can you be more innovative and bolt on and not wait for everything to come, right? It may be two, three years down the road before your imaging solution has built in native AI functions.

So it's either sit and weight mentality or if it's kind of getting outside of your box a little bit and moving away from that traditional mindset.

Bill Russell: Let's close with this. And this, one of those interesting one-off stories. I [00:16:00] hope it's not a one-off story, but it's an interesting story.

And you pulled it up. So I'm gonna ask you to comment on the story, but Emory Healthcare debuts first Apple powered hospital. And so it's Apple Systems from one end to the other. In a hundred bed hospital in Atlanta. That's a fairly diverse population mix. What do you take from this story?

What do you take from this effort?

Chris Jenkins: Yeah. I'll say it's somewhat innovative. To move away from a traditional Microsoft environments that everyone is used to using, whether they like it or not, right there, there's good and bad in that, it's costly. There's some vulnerabilities, but it's manageable, to make that shift

to a full Apple environment is, on the forefront of what any other organization would do. It's a little bit of a risk but I think there's a lot of value in doing that. But what, what do you have to do? You're probably gonna have to re-skill your IT team to some extent.

Those that support it. How do you provision de-provision those devices? How do you [00:17:00] manage them all the time? How do you teach your end users how to use them if they're used to a Microsoft environment, but what's the overall value if it integrates with Epic, which I know it does, and I think Epic is doing that if it reduces cyber vulnerabilities?

Today It does. Now in the future, will it continue to do that? You have to think through that, but it's pretty interesting that they would make that big shift. I'm really interested to see how it's going to continue and if that actually spawns others to go do something similar in that space.

Bill Russell: That is gonna be my question as well. I did listen to Alistair, Erskine is the CIO at Emory, and I heard him in front of a bunch of other CIOs and he gave eight reasons why Apple in the hospital setting and it was pretty compelling. Security was one sustainability was another.

He goes, each machine uses X amount less power.

Chris Jenkins: Yeah.

Bill Russell: Longevity of the machines. I know my wife's [00:18:00] iMac that's on her desk. If that were a pc, that thing would've been retired at least three or four years ago. Yeah. But the max, for whatever reason, maybe it's the walled garden of it or whatever.

They're able to last a lot longer than other devices. They're built, the aluminum and encasing and stuff, they're almost built for hospitals. they're really well designed. They're really powerful. Plus then you have the whole ecosystem. In this article they talked about using their iPhones, they talked about iPads at the bedside and in nurses stations, they talked about the Imax at the bedside and whatnot for documentation. There's an awful lot of stuff. I read this article I almost chuckled because. You read this article, it was all the executives taking a victory lap.

And I'm like you guys had so little to do with this. Do you know how big of a lift this was for the IT organization? Yeah,

Chris Jenkins: exactly.

Bill Russell: And not a single IT person is mentioned in the entire article. I'm like, wow. I know that's there's leadership for you.

Chris Jenkins: Yeah. Yeah. It's a lot of what you said there, the user interface, I think it [00:19:00] almost, I see an iPhone box in your background there, right?

So, well, that's the

Bill Russell: original iPhone box.

Chris Jenkins: Yeah. Yeah. Either iPhone or Android. I mean, people know those endpoints, but yeah, the user interface is another, it's pretty interesting. and I think you know, IT teams will probably be excited about it, to be honest with you.

Yeah, I mean, it's something different. I don't think it'll be hard for IT to understand how to do this. I don't know with the clinician, but if they do their user interface correctly and it's really just, point and click in, accessing that way, I think it'll be fine, but I really would love to go see it, to be honest with you.

, The other thing that

Bill Russell: Alistair brought up was eye strain. He said, the Mac monitors just have higher resolution. Yeah. And the nurses were really excited about the fact that they weren't getting as much eye strain, not as many headaches, that kind of stuff.

And you're like, wow, these are like things you don't even think about when you're designing a project like this. Yeah. Do you think other hospitals will do this or do you think this is gonna end up being one of those one-off stories?

Chris Jenkins: I think [00:20:00] Emory gonna get a lot of calls, let's put it that way.

I think other hospitals will start looking at this. I really do. I don't know that they'll go all in. There's always been pockets of, going back to when I wore a different hat bring your own device when I was in a different role a CTO role, it was always an iPad and that was connecting to Cerner back in the days.

Right. Which wasn't a very good experience, but even though it wasn't a good experience, the physicians wanted to bring their iPad in to connect with Cerner. So I do think there's probably be a little bit more of a drive for some organizations to do maybe pilots in some of this.

I don't know that they'll. Just carte blanche. Do a forklift upgrade. Oh. Yeah.

Bill Russell: No. I couldn't imagine anyone doing Yeah lift upgrade. That would be there's too much PC in the world. Yeah way too much and knowing what I know about just the technology that had to go into the back end of this it's not a small lift.

No. In order to do this this also ended up being sort of a greenfield situation in that everything in that hospital needed to be replaced. It was so old and [00:21:00] so outdated. Yeah. That they were like, look we have to replace it all anyway. We could just run through there and put PCs in there and they just saw an opportunity to change the game.

Well. It turns out that, user community, as you read in the article, is just ecstatic with the results. I think they will get a lot of action. I think a lot of people will come through and look at it. It will only be the most innovative organizations that do it. So, as much as I am a huge fan of what they have done here, and I think they have broken through a barrier.

more and more I think the CIO is platforms, standards. they're not gonna take on something like this that's gonna be seen as an outlier unless it has a material impact. I believe it does have a material impact, but I'm not sure they're gonna see it the same way I see it.

So, Good points. It's kinda awesome. Well, Chris, thank you for talking about the news. Yeah. With me and for muddling through on this Tuesday morning post Memorial Day. Still trying to get my [00:22:00] voice and my thoughts back. It's gonna be fun. Hey, thanks again and look forward to talking to you again.

Sounds good. Thanks Bill.

Thanks for listening to Newstay. There's a lot happening in our industry and while Newstay covers interesting stuff, another way to stay informed is by subscribing to our daily insights email, which delivers Expertly curated health IT news straight to your inbox. Sign up at thisweekealth. com slash news.

Thanks for listening. That's all for now