Welcome to PathPulse: The Digital Diagnostic Podcast! In a world rapidly embracing digital transformation, the field of pathology is no exception. We’re on a mission to showcase the pioneers, innovators, and forward-thinkers who are not just observing but actively shaping the day-to-day use of digital pathology.
In this episode, join our host, Andy Ivie—board member of the Digital Diagnostic Summit and Lumea’s Chief Innovation Officer—as he sits down with Imogen Fitt, a principal analyst at Signify Research and a leading voice in the healthcare technology market.
Get ready as we dive deep into what pathologists really want from digital pathology technologies. From the initial hurdles of adoption to the cutting-edge potential of AI, Imogen shares her unique perspective on the disconnects, the desires, and the future of digital pathology from the viewpoint of those on the front lines. You’ll hear about the critical need for seamless integration, the evolving role of AI beyond the hype, and the often-overlooked importance of robust support and interoperability.
Tune in to discover what truly matters to pathologists as they navigate the transition from glass to digital, and gain unparalleled insights into how the industry can better meet their needs to unlock the full potential of digital diagnostics.
Listen here:
Transcript:
Andy Ivie: All right. 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 Andy Ivie. I sit on the board of the Digital Diagnostic Summit, and I am Lumea’s Chief Innovation Officer. And in this episode, we’re exploring the theme “Digital Pathology Tech: What do pathologists actually want?” with Imogen Fitt. Imogen, welcome. I’ll give you a chance to introduce yourself and share a little bit about your background. And then I have a surprise question for you. So let’s get started.
Imogen: Well, hello listeners. My name is Imogen Fitt. I am a principal analyst at a company called Signify Research, which is a market research firm that focuses exclusively on healthcare technologies. So I’ve been covering the digital pathology market, speaking to vendors about its evolution, what they’re seeing in terms of trends driving growth, until about 2021 now, and it’s been great to see the progress that we’ve made so far. And I post a lot on LinkedIn in terms of insights from the industry and where it’s going.
Andy Ivie: Great. Thanks for sharing that. So the surprise question is: Give us a little color about yourself, not just your work background but, like, what makes Imogen the cool person we all know?
Imogen:I don’t think my colleagues would describe me as cool. But my nickname can be bubbles because I have quite a bubbly personality.
I genuinely just love health tech. I love seeing these innovations come to life. My background is a biomedical science degree from the University of Warwick. I used to look down a microscope myself a lot as part of the learning and the research for that role.
However, I also remember getting quite frustrated with learning to use the microscope and thinking, “Why haven’t we moved past this yet? Why is this so old?”
When I graduated and found my first job, was Signify Research, I was absolutely thrilled to learn about digital pathology. And I was completely on board with it then and really excited to learn about how it could be adopted within clinical settings specifically.
So I don’t know, I think I’m quite boring. But I guess what makes me a little bit different is I’m quite happy, bubbly, passionate, ready to talk to anybody about the market.
Andy Ivie: I think Bubbly is maybe two or three notches above cool, so kudos for that. And I think everyone is right: Imogen, the bubbliest person in pathology
So we’ll leave a little bit there for future podcasts. We’ll dive more into your bubbly life. But let’s get into some of these topical questions.
First, let’s talk a little bit about the disconnect and the concerns you’ve heard from pathologists who are getting started with digital pathology or have been digital pathologists for a little while.
Like, what are they expecting that’s not being delivered, and what are their, like, pain points?
Imogen: So there’s a couple of things and they relate separately to digital pathology as a product and then also the change management process in introducing digital pathology into workflows because it’s not as simple as just selling the product and then suddenly the pathologist adopts it and all is hunky dory.
And I’ll touch on the latter first because I think it’s something that is really important to pathologists at the moment, especially because we’re at such a nascent stage in the market. For pathologists, this is obviously such a massive change from their day-to-day workflows that you can’t just expect them to adopt digital pathology and…sell the product and essentially just leave them there.
What I’ve found and what’s really been communicated to me is that that initial vendor onboarding and training process is something that’s often the differentiator between going between vendor A or vendor B when you go through that selection process. There’s almost a sort of a hand to hold element that needs to be done.
You need to be guided through the process, guided through the software, how it works, and especially through those initial first months where you might be scanning sort of a minimum of cases, you need to have a dedicated team of service, like maintenance and support, that’s they’re really there to be on the other end of the phone at all times to answer questions when they have it, to be able to sort of solution solve when something does come up. Because something’s always going to come up when you’re integrating into a new environment.
So I’d say that’s the most major thing outside of the product itself that is the gap that pathologists are really keen to sort of build the bridge between today. And then also, and this ties into it, there’s also an element and a need to really integrate digital pathology within the wider healthcare IT environment and the wider research environments as well. So when we talk about interoperability, it needs, your digital pathology, your IMS software, needs to be able to speak to the LIS, speak to the EHR as required.
Because if you’re inputting and already using a bit of software, what you don’t want to have to do, the last thing you want to have to do, is sort of input data into the LIS as you’ve reported it and then go ahead and have a new pop-up window that, automatically, you have to then input it all in again. That’s something that I find is quite a prominent concern today and quite a prominent frustration, because things often don’t work as planned and it often takes a lot longer to set up those smooth and tight integrations that are really required for just an easy workflow.
Andy Ivie: And no matter where you start, there’s definitely a goal of a very smooth interaction between the diagnosis and all of the case information and the images themselves so that the pathologists can continue to operate very efficiently just like they do currently or even more efficiently.
So I know AI is maybe most pathologists’ entry point into digital pathology or like that carrot hanging out there that makes them want to go digital. I’ve also been at a lot of conferences, like you have, where we’ve seen each other and I’ve been in some presentations where people ask the audience who’s using AI clinically and who wants to use AI clinically and who would use AI clinically if it was free, who would use AI clinically if it cost money.
And usually it’s a pretty, kind of shockingly small number of people. So what do you think the AI can realistically do in clinical pathology diagnostic workflows right now? And what do you think is out there on the horizon that people should be looking forward to?
Imogen: So when I think about AI, I kind of group it into different brackets because a co-pilot, which is kind of on the more sophisticated end of the scale, does a lot of things that you need it to, it’s very complex. It’s very different to a simple tool that counts cells in an image. And often it’s the simple tools, the cell counting, which can be quite a great low-hanging fruit because it immediately solves a significant pain point for pathologists. Because they don’t want to have to sit there and go one, two, three, four, five, six… if you see what I mean. For the case, however, it could be hard to commercialize and make an effective, create an effective business out of the latter because you’re not going to be charging quite a lot for a solution that does something as simple as that. What I’m seeing is, at the minute, a lot of interest in QC tools, and that’s brought about by going back to some of the pain points that we mentioned.
Digitizing slides is often sort of a step that’s far, far down the line. And if you’ve not produced the slide itself properly, or there’s a bit of a blur or something’s not quite right, it can get all the way to the pathologist sometimes in the IMS for review before anybody picks that up. So that can be quite annoying to have to sit there and constantly flick through slides and go, okay, well, that’s not working. I need to resend that for rescanning. I need to do this. I need to do that.
So what we’re seeing is a lot of scanner vendors recently partnering with QC, or developing themselves QC or QA software, which is designed to nip that issue in the bud and quickly address that need.
Now, those sorts of solutions will be kind of additive in terms of cost, and there’ll be a delicate conversation between providers and what, essentially, they’re willing to pay because there is a limited pot of money. And I think that’s part of the reason actually that AI is quite limited today, because the scanners are expensive. You want to then invest in the IMS which allows you to coordinate your work across large geographies or a different number of sites. And then you have the conversation about AI and storage and all those sorts of other things.
So quite often, unless something’s planned quite…detailed upfront, you often find that the scanners themselves take up a lot of the budget and providers have to make do with the scanner sort of bundled software for a little while. We’re seeing that slowly change, more more adoption.
But going back to the AI question, QC and QA can be quite low-ranking for you, but again, it isn’t something that you can really charge for because it doesn’t actually provide that much value back. It’s still doing something that the pathologists can do themselves. What we’re seeing more and more vendors lean towards now is prognostic AI, which is able to identify genomic biomarkers. And those kinds of applications are really exciting because that’s something that the pathologists wouldn’t have been able to have done beforehand, but which could quite significantly alter the workflow and the timeline for a patient to be able to get a diagnosis.
Because if there’s an algorithm there that flags that maybe it’s appropriate to do some genetic testing, even if we do the genetic testing because those algorithms might not replace the genomics testing itself, it means that it’s flagged that much earlier in the process and that you can get to an endpoint and to the treatment pathway a lot quicker than usual.
So I guess it depends. There’s QC AI, there’s simple stuff like cell counting, there’s your diagnostic and triaging, there’s also your predictive prognostics. And then there’s copilots. There’s so many different use cases for it today that I think it can be argued a little bit as well that it’s hard for pathologists and users to kind of break through the noise and assess quite quickly which solutions are going to be most applicable and useful in their unique setting right away.
Andy Ivie: I think the quality message of AI is one that resonates strongly with people. And that providing more tools for quality control and quality assurance, both in the slide and in the diagnosis and in the prognostics, could be pretty effective. You still have to deal with the economics of it, which I think everyone’s still trying to figure out how not to be the one who maybe benefits the least for it, but you’re the one who has to pay for it. And the lack of AI reimbursement is a real challenge.
And getting those tools in the places where they would provide the most benefit to patients. And so I’m really hopeful that we have a breakthrough in realizing the value of these tools, putting them in the right place, and covering their costs so that patients can have higher quality care and higher quality results.
And so, yeah, I think you’re spot on with the ways that people can use it now and some of the cool new things that people are doing with it. I think it’s always, at least in the US and even abroad at some level, you at least have to make a value pitch of how it’s going to improve things. And then finding the right way to get it paid for is important.
Imogen: And kind of all hope isn’t lost because I know that there’s some PLA codes for some algorithms in the US market. I think it’s Tempus, Arterra and Imagine, as well as PreciseDX that have achieved them. So at least there’s reimbursement for those type of algorithms internationally as well. Korea has brought in, or South Korea has brought in, some reimbursement for AI.
But I think the important thing to remember as well, within those circumstances that other people kind of get excited about, is that you still need to make the business case for the entire digitization process.
Andy Ivie: Yeah, definitely the coolest parts of digital pathology are the things that you can’t do on glass or the things that are really expensive and take forever to do on glass. And so every time there’s a new tool in AI or on the viewer side or anywhere in that process that is either incredibly cumbersome or actually impossible to do on glass, I think that’s where, like, the magic happens.
We want to make the day-to-day pathologist experience as good as possible and try to improve things over the glass workflow. But when you can do something magical that you can’t do with a physical glass slide, then things start really popping up. So that’s when I see pathologists get bubbly. Maybe not as bubbly as you, but a little bubbly.
But yeah, any advice? I work with a digital pathology company and so I see some of the wins and the challenges of digital pathology firsthand. You work with and talk to a lot of people in industry and the end users of products like Lumea’s. What do you think is the thing people want the most that they’re not getting or the thing that they’re getting that they like the least, where can we improve the digital pathology experience for users?
Imogen: So I think it’s important to be mindful that the market is bifurcating between those who have experience with digital pathology and are maybe not using it to its full potential and those that are just embarking on that journey.
And those that are just embarking, it’s very much a stepwise process where they need a partner as opposed to a vendor to sell them a solution. They want advice, they want things like the IMS software and vendors being asked about, how should I treat storage? And a lot of the conversation that I’ve seen in the industry recently is you need to get your IT department involved so that they can offer you some more advice on that.
But even the vendor sort of saying, “You need to do that” is something that’s incredibly helpful for the lab if they’ve not yet involved it as part of an enterprise imaging strategy or anything of the sort, if they’re just going at it alone. Because sooner or later, they will benefit from that integration and being able to lean on the experience of other healthcare stakeholders within the environment that they’re in.
In terms of those that are much more sophisticated and maybe a little bit in the middling of their journey, they’re having scanned slides a little bit, they’re looking to scale up workflows.
They desperately want to be able to, it comes down to interoperability and that’s why it’s a buzzword in the industry at the minute. They want to be able to partner with vendors that are going to be open to employing different solutions from third party providers. And that includes AI. That’s why there’s been many AI partnerships with, we call them orchestration platforms. You might call them like an AI marketplace or something like that.
Providers themselves don’t want to partner with just a single vendor. They would rather streamline the supply chain and manage that relationship through just one party, essentially, so that they can purchase, they can do everything that they need from one house. And they’re going to do that whether or not the pathology vendor wants them to or not. They’ll just find another way to do it.
If you look then, so that’s part of the reason as well that they want, they’re very keen on adopting AI, but they want to do it in a way that’s more beneficial to them. Because a lot of business is generated direct to providers today, but they do, yeah, they do want to be able to do that through third parties and integrate with as many different third party AIs as possible. It also comes down to… one I’m hearing a lot about is integration with VNAs. So vendor neutral archives are already situated within radiology departments, and they’ve been kind of growing into cardiology and it used to be all specialized, but now it’s coming under this enterprise imaging umbrella.
Enterprise imaging, which has traditionally always been radiology, is now cardiology plus maybe ophthalmology, is increasingly wanting to sort of take in digital pathology. And from a storage perspective, that can be quite… attractive because it enables you to sequester additional budget from outside the lab itself in order to leverage shared, in-house resources.
So I was speaking to a sort of a big network or the head of a big network in the UK the other day and I asked him, you know, what do want me to tell the vendors? What do want me to speak about and they were like I want to be able to view my digital pathology images in the VNA and in the in the viewers and I want better integration with radiology. I was like that’s interesting because, and bear in mind, their setup was very sophisticated. So they were already incorporating QC, QA software. They had sort of a robotic setup for the samples around the lab. They’ve been using the scanners for years. And for them, it was much more important to have a holistic hospital IT network that worked cohesively all together. There’s something called the….
There’s a hospital in South Korea which recently achieved, I think it’s Dioram? I’ve probably got that wrong. But there’s a society for imaging informatics, SIIM, they all kind of work together to try and guide hospitals in how to set up their general healthcare IT networks so that they work interoperably.
And the hospital, which I think was Samsung in South Korea, was the first in the world to achieve that. But they were so far along in their digital pathology journey that they’d moved beyond just digital pathology as it works. They were like, well how does digital pathology work in wider healthcare? So as you go along, you get different stages of maturity and I think the priority shifts between them and there’s so many nuances in between that as well but hopefully that was helpful.
Andy Ivie: Yeah, yeah, I think you really drove home the point around interoperability with whatever system the pathologists in the lab want to work with.
I think that covers, you know, different scanner vendors in a single lab, because they may have specific use cases and capabilities, to a variety of AI solutions and a variety of storage solutions for archival and for hot storage. So I think orchestrating all of that is a big task and something that I think the industry is getting better at.
I’m curious, do you have any kind of anecdotal stories of groups that have done a really good job with integration and the timelines that that took and maybe groups that really struggled and what they reported as their pain or failure points and how, maybe, we could avoid those in the future?
Imogen: So there are a few examples of great digital pathology implementations. I’m not aware of the timelines, more because they were up and running when I first heard about them. So there’s the hospital in South Korea, the Samsung Hospital, that I know has done really well. There’s also one in like a network in Northern Ireland that’s producing petabytes of data. And they also looked at it as more of a…Northern Ireland has been sorting out of its digital pathology strategy for a while.
Timelines tend to take years to do it properly, because, especially if you’re introducing digital to the public for first time, if you look at the UK market, which is arguably one of the largest in terms of revenue generation for many of the vendors recently in recent years relative to, you know, pathologist proportion and size, la di da di da.
Even though many labs over here have access to a scanner, many of them are still only scanning between, you know, 10%, 20%, 30% of slides maximum because they’ve bought and sort of instigated all these scanners and installed them and it’s taken them a while to get the pathologist workforce on board to decide exactly how much of the workflow they want to digitize.
And then before they even sort of consider the question of, “Do we store it or do we start applying AI?” they’re worrying about so much in terms of little things like “Are the barcodes correct?” Because there’s challenges around different vendors having different barcodes, scanners formats, we haven’t even spoken about DICOM and obviously all of the proprietary formats that exist and have made things, I think, a little bit more difficult historically for pathologists that want that interoperability and want to be able to change between different vendors.
So in terms of institutions where you can sort of look to them for examples of how to do digital pathology very well, I think there needs to be more around that, more case studies published, but I also think organizations like the DPA, the ECP, different, there are emerging bodies that are looking at themselves as authorities to help guide that process as well.
I’m sure Lumea is also part of a few of them and the Digital Diagnostic Summit will be a sort of platform where pathologists can gather and talk about their issues and really sort of use it to leapfrog some of the common challenges in adoption I should think.
Andy Ivie: Yeah, that’s a great point. I think there’s been an evolution in the way people share and the groups that share their stories. I remember hearing a lot of academic institutions in the early days talking about how they adopted digital pathology for tumor boards or for education.
I think, more and more now, I hear groups that are going digital in a 99% kind of way. Or even that 10 or 20 % is very different than just tumor boards or just education. It’s actually operationalizing digital pathology, becoming a digital pathologist, using your microscope less and less frequently to the point where, you know, your default behavior is being a digital pathologist. And I think that’s a really cool trend, one that I’m excited to see grow in the US and internationally.
I think digital pathology is the future. I think if I had a biopsy, I would want it to be read out digitally. I’d want it to have all of the most, like, advanced tools, the cool new stuff, you know, I think that’s what patients want. They want to have the best care possible and to make sure that their health providers are at the forefront of all that technology.
And so I think the patients are going to drive us in this direction because they want the best care. They want to be taken care of. They want their loved ones to be taken care of. And so digital pathology is here to stay and just a matter of time, I think, as we grow.
Imogen: So that brings me to a really interesting point. I don’t know whether you saw Google and Endeavor Health’s announcement, but the press release really struck me because they were using digital pathology as a tool to bring, or they were planning to use digital pathology as a tool to bring pathologists closer to the patient. Because obviously, historically, you don’t really meet with pathologists. When you talk about, know, the patient wants the best technology and they want the best support possible that they can get.
Now, in radiology, that’s relatively easy to sort of also market your products to the patient themselves because you have outpatient imaging centers, you have quite a lot of decisions driven by the patient themselves, whereas in pathology, I think it’s arguably less of a factor today, but it is something interesting to consider with strategies moving forward.
Is there sort of a market for speaking directly to the patient themselves and educating them and saying, “Do you want to request digital pathology imaging?” and those sorts of things. If memory serves, there’s something around the South Korean reimbursement for AI—I think it has to be driven by the patient as well themselves and they have to be notified and involved in the process. So it’s something that just struck me based on what you said, because I think that’s quite an interesting point to make.
Andy Ivie: Yeah, and I think digital pathology will unlock more equality of access in areas of developed countries that have kind of less access to high quality pathology, but also in developing countries where, you know, I hear stories of months to move tissue or glass around to somebody that can give an opinion on it and waiting months to maybe get a better answer.
You know, that’s the kind of logistical improvement digital pathology can bring. Same for countries that kind of leapfrog the telephone line straight to smartphones. I think a lot of countries will leapfrog the glass histology infrastructure to a digital pathology future where you can very quickly access high-quality pathologists and be able to do these AI tests from wherever you are.
Because if you have a digital image, you can do much more with it much more quickly than you can with a glass slide or with a piece of block of tissue. There’s a lot more options. But, anyway, you know, you and me, Imogen, we could go forever on this stuff.
We’re gonna have to make sure we have you back sometime so we can dive in deeper. But glad to have the most bubbly person in pathology on our podcast.
Imogen: I don’t want that be a thing. We can’t make that be a thing. That can’t be a label.
Andy Ivie: Okay, I’ll say “top five bubbliest people in pathology”. We love making top five and top 10 lists. But number one in my book, Imogen, thanks for coming on. Thanks for sharing your insights with us and your broad experience with pathology, especially digital pathology.
Looking forward to sharing this with the world so that people can hear your great insights. And I guess to all of our listeners, thanks for tuning in. Join us next month for another episode of PathPulse, the Digital Diagnostic Podcast. Thanks again, Imogen.
Imogen: Thanks for having me.
