Welcome to PathPulse, the Digital Diagnostic Podcast! In this episode, your host, James Thackeray, speaks with a true trailblazer in the field: Dr. Orly Ardon. Get ready to dive deep into the fascinating and complex journey of adopting digital pathology in the US hospital market.
Dr. Ardon, Director of Digital Pathology Operations at Memorial Sloan Kettering Cancer Center (MSK), shares her unparalleled insights from leading one of the world’s largest digital pathology operations. From the initial hurdles of building an entirely new digital infrastructure to the incredible horizons of AI and improved patient care, this conversation explores the real-world complexities and immense potential of this transformative technology. Whether you’re a seasoned professional or new to the digital pathology landscape, you’ll gain invaluable perspectives on how institutions are navigating this crucial shift.
Listen now:
Transcript:
James Thackeray: Welcome to PathPulse, the digital diagnostic podcast. This podcast showcases pioneers, innovators, and forward-thinking individuals within the digital pathology arena who are making a difference in day-to-day use. I’m James Thackeray. And in today’s episode, we’re tackling a topic that is central to the future of our field: from hurdles to horizons, the complex journey of adopting digital pathology in the US hospital market.
We are honored, and I’m so excited to have a friend of mine, and colleague, and just someone who I actually think is a superstar in the digital pathology world, Dr. Orly Ardon, the Director of Digital Pathology Operations at Memorial Sloan Kettering Cancer Center, where she leads one of the largest digital pathology operations in the world. I’ve seen it, it’s incredibly impressive. We’re really excited to have you here, Dr. Ardon, today with us on the podcast.
Orly Ardon: Thank you. Thanks so much. Thanks for having me.
James Thackeray: Yeah, well, as opposed to – why don’t you start with just an intro, maybe give us a little bit of your background, and then the role that you’re playing at MSK and have been playing as this digital pathology implementation, and kind of everything that you’ve built there has played out.
Orly Ardon: Sure, so I’m not a pathologist. Let’s start with that. I have a bachelor’s, master’s, and PhD in Microbiology from the Hebrew University of Jerusalem. I came to the US, to Utah, and I spent some years doing a post-doctorate there, and then I started working for ARUP Laboratories, mostly doing research and development.
And at some point, I got into the digital pathology world with a mentor with Dr. Mohammed Salam over there. So at MSK, I’m working on trying to really get this huge operation that’s pretty, it’s the largest, probably one of the largest in the country for sure, but it’s really comprised of many different individuals, many different departments, who are doing a large job. It’s not just me doing everything. And this work was also built on pretty much the shoulders of giants who started working on this in the early 2000s when digital pathology was still in its infancy.
So we are now scanning about 6,000 slides a day. We have 38 high-throughput whole slide imaging systems, meaning 38 high-throughput scanners, not all of the same make and the same vendor, so it’s a very complex environment. In addition, we have smaller models that we use for dedicated tasks. Our goal, and we’re very proud that we got to this point, it was to really find a solution to digitize every pathology asset that we have here in the department.
And Memorial Sloan Kettering, it’s a cancer center. We are a very complex center in just the nature of the specimens that we get here from all over the world, and very rare cancers. And we have some needs that smaller places or reference laboratories don’t necessarily have. So…
So in 2020, we were able to validate a remote sign out and test for digital pathology, which really kicked off this clinical use for digital pathology. So from 2020 onwards, we’re able to start scanning prospectively and allow pathologists to sign out without having the glass slide next to them, meaning they could work remotely, they can work here.
The reality is that we still have glass slides. So even though we are now at the point in which we digitize 100% of the slides that the laboratory is producing, for most of the services, we still distribute the glass slides. So we had this phased approach of increasing our scanning, we find solutions, but we’re still at the point in which we have a slow adoption phase for digital pathology.
At the same time, we are also providing solutions for our research. This is a research institution. We have trainees. We have lots of education initiatives, and we try to share the work that we do. We offer courses here. We offer lots of collaboration opportunities. We work with other institutions, and our goal is really to get as much digital pathology to make it more easy for other institutions to adopt and to use.
I’m also working with the, I’m on the board of the Digital Pathology Association, I’m on some committees there, and the goal is for me is really to see more innovation in pathology, to see more adoption of new technologies, but also to look at the economics behind it to see how we can deliver some revenue, some return on investment to our departments.
Healthcare is always in need of more funding. And some of the research that I do in my role is dedicated towards that, really looking at the healthcare economics and how to really improve the way we operate. So that was a long introduction.
James Thackeray: That is perfect, and it’s a great overview of where I think we want to kind of dive into. I think your experience in this full from concept all the way into full-fledged digitization of all your slides, and there’s still phases to come. I understand that, but you reference the complexity, and I think some of this is specific to MSK, and a large institution, an academic institution with a research side of it, adds to some of the complexity that you deal with.
So let’s start with that, the complexity side of it. If you were to highlight one or two of the major hurdles in getting into this, and maybe if you would, and I’ll kind of chime in too, comment on where some of those challenges historically have been and where, maybe, they’ve lessened over time.
Maybe we’ve gotten better at doing some of these things over time as well, because you guys were so early into this. But start with kind of highlighting the complexities that you faced early on in this life cycle.
Orly Ardon: Definitely. Yeah, we are definitely, you know, we are early adopters. So that means that we make mistakes. We have no guidebooks. We have no publications. And sometimes we work along with our vendors to find solutions that would be available for the later adopters to use. So there are definitely a few issues that have plagued us from the beginning.
I think the complexity that you mentioned is just the nature of working in healthcare, making sure that you’re working in a patient-centric environment. You want to make sure that whatever you’re doing it doesn’t pose any risk to the patient. Meaning, we want to make sure that we have the right image of the right patient at the right time to the right pathologist. Let’s start with that.
How do you manage to get those images, those large gigabytes, sometimes two-gigabyte images, to the place, to that specific location, where they would be available in the other systems of the hospitals? We have the laboratory information system, we have different viewers, we have different servers, networks.
I mean, the complexity of the healthcare environment, the digital healthcare environment, is such that every time you try to disrupt it, add something new, you have to really think about all those other details.
Sometimes it’s easier, especially if you work with just one system and you can get a lot of vendor support to just plug and play, but the reality, it’s never a plug and play. There is a lot of preparation work, a lot of scoping what we’re going to do, it can take years before you actually take this idea of adding a digital technology.
And from the moment you take this idea of the digital technology until you really get to a point in which you can use it at the hospital environment. But it’s getting easier. It’s getting much faster.
And the truth is that we are more complex because of the different systems that we have. So we have to find interoperability, something that the early digital pathology field did not have. Now, it’s on everybody’s mind. It’s the reality in which we start seeing more vendors working together, hardware, software. Obviously, I think it’s pretty obvious that without having that, you cannot really market your tools. You cannot really have a good environment in which we are able to use different systems for different needs.
So I do think that we got to a point, you know, in the early days of work here, we had to develop our own digital pathology viewer because one was not available and now every image viewer that you get, you can get commercially, is interoperable.
Another major issue that we have is the file types that are always a cause for every update that we do and we have to always invest some time thinking about any change that we do. It seems like we have those butterfly effects if you don’t plan accordingly.
So it does take a village. I always get asked about hurdles, and I think the biggest advice I always give is that you have to really not just rely on the pathologist or not just rely on the technicians who are scanning, but you have to have this ecosystem of individuals with the right technical background.
We work with the IT department, anyone from servers, networks, it’s such a huge field. From the LIS people, different teams who support the technical part, the hospital, the information systems, it’s a huge ecosystem.
Everyone should be trained, everyone should understand this architecture of the digital pathology operation because you know this is relatively new and we inserted ourselves into a very restrictive pathology operation, but without having this support from all the other stakeholders, we just cannot operate.
James Thackeray: That’s so great. The complexity of integrations, storage, you kind of mentioned the server aspect, I mean, some elements that you probably originally wouldn’t think about that you guys have dealt with early. And I love your summary of it takes a village. And so if you’re at a hospital system out there, a hospital that’s looking at this, getting the key stakeholders involved early to kind of understand—one, the why of it, but the complexity of the how, is probably pretty critical.
That’s a question I had is, was MSK as an institution, I mean, they’re always leading in research anyway. How was that process of getting key stakeholders involved in going forward with digital pathology in general? Give us a little background on how that went down, because that’s pretty significant, especially seeing when you guys started into this journey.
Orly Ardon: Yeah, so I cannot take credit for anything that happened before 2020 because this is when I joined. But the story is that we had some amazing visionaries here who saw the benefit of the technology at a time when it was still an idea with very few proof of concepts.
And it did take investments in resources, and we had leadership who understood where those digital technologies would take us at a time when there was still no AI in pathology, really. There was still no—this digital ecosystems and solutions and nobody understood the amazing opportunities that there are in digital technology.
So the fact that we were able to get these phased investments in technologies and this is something that is, I need to emphasize this, didn’t happen overnight, it took many years of building technologies, capabilities, setting up teams, trainings. I mean, it’s a massive operation, and you do need to prepare for it.
So I think one of the most important things is just to constantly talk about the benefits and look at the other things that are happening around the hospital, which include digital technologies. Because let’s face it, without the pathology digital images, you really cannot have an entire digital operation at the hospital.
So you need to start somewhere—you need to start investing in those technologies in order to get pathology into this reality of electronic records. We just cannot continue to have this manual glass slide operation when everything around you is digital.
James Thackeray: I love it. So now we’ve kind of talked about the implementation process, and I guess I’m reminding our listeners, too, that I think what’s exciting about this is we’re talking about kind of the extreme side of it.
You think you think of the level that MSK is at and all the different systems you’re integrating with, I’m also trying to say, okay, how do we, I I think for a hospital, individual hospital, it’s going to be a smoother sailing because of the pioneering that somebody like MSK and you guys have done there. It’s gonna help others be able to adopt more readily.
Now, as you look at this, you’ve got full implementation. You’ve talked about phased approaches into it. If you look at it today where MSK is at, one, so I have a, how much is research-based use right now, and how much is clinical sign out, like digital pathology, clinical, primary diagnosis sign out, and where do see that next phase going? Is that fair?
Orly Ardon: Sure, yeah. So, you know, the first thing we do is we talk about clinical utility, the clinical use of those digital images. There is really no justification for us not to do all of that if it wasn’t for the clinical need.
So you set up this ecosystem, the infrastructure, the base for what you need, and now you’ve got an amazing opportunity to benefit the clinicians, the pathologists, everyone who’s working on patient care can benefit from those digital images.
Along those lines, you’re still having a huge research repository that slowly gets built up and can be used for research. So I can tell you that—I don’t have exact numbers about how many pathologists sign out exclusively digitally. Most of them open the images, some of them sign out, others choose to look at the image and turn around and use the microscope. We still have microscopes in pathologist offices. Most of them did not give that up.
All the data is being used for research. We use the data for education purposes. And we do have, you know, once you have an image, you can use, this is the beauty of the digital reality, you can use it. We hope to increase the digital sign out so that in the future we will not have to deal with the added expense of this hybrid workflow in which we have both the glass and the digital image, but at the same time, we know that it’s a matter of time.
There are things that, you know, there are lots of human aspects about changing the way you do the trade and not everyone is ready for this digital transformation. So we try to make it as smooth as possible, but the future will look very different than what it is now today.
James Thackeray: I love that. And I think that’s such a great approach, this phased approach. And it’s not, “You have to adopt tomorrow,” because I think you’d have the resistance, but so you’d have the wrong mindset going into a new technology where when you let pathologists kind of work themselves into it and understand it better, a great, great approach for sure.
Talk to us about your role or how you foresee your role in educating outside of even MSK as others are looking at going digital. Based off of your experience, I know that you’re passionate about helping others kind of go down this path and learning from things that you guys have already been through. Talk to us a little bit more about that side of it.
Orly Ardon: Sure, and I think it’s just our duty, as people who have already done it. We are scientists, this is what we do, we want to share this knowledge. And the truth is it’s a bit selfish because the beauty of digital pathology is that if more institutions go digital, it benefits everybody.
Right now the market for the medical devices for the scanners is pretty small. The more laboratories, the more healthcare systems get into digital pathology, we’re going to be able to expand that, have more image sharing.
We’re going to be able to send for consultations. Places that don’t have the pathologists around, but they have a way to get a simple digital scanner, a device that allows them to send a digital image elsewhere while the patient is waiting for treatment.
So for us as a cancer center, the ability to help promote that and see global health care really benefit from those innovations is simply priceless. This is what we do. And the truth is, we want others to learn from our mistakes too. You sometimes they talk about failures, sometimes they talk about challenges. And I know it’s not as sexy as talking about innovation, but we have to understand that if we don’t share those failures, others are going to repeat the same path that we took.
And if we can help others save some time, obviously, their resources, by sharing this and you know, it’s a lot of effort to bring people in—we do a course every year here—but we want others to see what we do. We want them to be comfortable with digital pathology operations. We can do this. So I think it’s really a responsibility for all of us to educate, to publish, to talk about it and to try to move this industry forward as much as we can.
James Thackeray: So great. And you feel your passion for that. I’ve always, since I’ve known you, known that that’s a big aspect of this. Similar to a common acquaintance we have, obviously more than acquaintance to me, but our founder at Lumea, Dr. Matt Leavitt, whose vision has always been that digital pathology can help where we have little diagnostics in parts of the world.
We think that that’s always in a third world country, or we label it that way, but it’s really globally, we have a lot that we can do on the diagnostic front. And so I know his passion is similar to yours and how do we utilize digital pathology to get diagnostics available out there and all sorts of things, right? I mean, second opinions from leading institutions around the world that can take as much as 30 minutes through an integration as opposed to, who knows, maybe no access at all to the world leaders in a specific specialty. So I don’t mean to go off on it. I just know that that’s a common patch between our founder and you.
Orly Ardon: You know, and knowing the digital pathology community, we actually are a very cohesive group of individuals who are very passionate about this. So if you come to those digital pathology meetings, I know your meeting, you see this energy.
A lot of people who want to move forward and do things, and there is so much potential. Again, not everyone has the resources that we have in the US, definitely not in a place like New York City that has so many academic medical centers, especially MSK, but you’re talking about expertise, there are areas in the US that don’t have the healthcare that they should be getting, so digital can really solve some of those issues, but we first need to put the infrastructure there, and it’s always a barrier.
James Thackeray: Agreed, and there’s so much that you’ve learned along the way. Okay, in summary, since we kind of started with, you know, if this is more for hospital setting type, you know, potential integrations and implementations here, what’s your one minute summary of why a hospital, or at least, why you believe a hospital might look at digital pathology as a priority at whatever timeline they maybe deem necessary, but as a priority for the future.
Orly Ardon: Sure, you know, and I do believe that most hospitals are talking about it or starting to explore not just the, it used to be just the large academic medical centers, but now we see more smaller hospitals. And I do think just from the interest that they hear, there was more plans and it’s really just a matter of resources at this point.
And the why, the why is that this is better patient care. When we hear about lost glass slides or broken slides that are not usable and it’s a patient specimen, when you have a digital image, it does not get lost. This is just one example of the fact that you can get rapid diagnosis because you can use an expert who is not in the room with you with that glass slide. You can call someone, you can do so much when you have a digital image.
And then there’s the promise of AI. If we thought that it was a novelty 10 years ago, right now, we know the technologies are there. They may not get onboarded as fast as some of the people who develop those would like to see because there’s lots of hurdles in healthcare.
But at some point there will be AI tools that will relieve us of some of the mundane tasks, the ones that can really solve issues of workforce and aging populations. And the fact that we just don’t have enough pathologists, we don’t have enough histotechs, we don’t have enough staff in pathology. And if you can use digital technologies to just enable the operation and the ability to provide pathology services for the near future, I think it’s priceless.
James Thackeray: Beautiful summary. We should just maybe go right to the summary because that is the rationale for institutions to look at going digital if they haven’t already. And I agree that they, most are looking at it right now.
Orly Ardon: Most are, yeah.
James Thackeray: Well, thank you. You’re so great. We’re so grateful for all the work that you’re doing. We look forward to seeing you at the Digital Diagnostic Summit in September. Little plug, but also, you’re gonna be there and speaking, we’re grateful to have you. Thank you for joining us.
Orly Ardon: Yeah, thank you.
