Global Digital Pathology Task Force
01:13:33 Share

Welcome back to a very special episode of Beyond the Scope. We have the honor of welcoming three presidents of three respective digital pathology societies on three different continents. Dr. Junya Fukoka is the president of the Japanese Society of Digital Pathology, and he is also a professor of pathology at Nagasaki University and a lung pathology specialist with a cross-appointment at Kamada Medical Center in Japan. Norman Zerba is

is the president of the European Society of Digital and Integrative Pathology. He is also a computer scientist at the Institute of Pathology at Charity University Hospital in Berlin, Germany. Esther Abels is the current president of the Digital Pathology Association. She is also the chief clinical and regulatory officer at Visiopharm.

I am David Tolman, an entrepreneur and engineer in the digital pathology space. I'm also co-host and producer of Beyond the Scope. This special episode is hosted by Giovanni Lujan, a pathologist and associate director of digital and computational pathology at The Ohio State University. What ensues is a fascinating panel discussion about the state of global digital pathology adoption. What are the use cases on each continent?

How fast are we moving? What standards are required to move forward and create a global collaboration? What are the regulatory and ethical considerations dealing with patient data? It's exciting to hear answers to these questions from the three presidents. Beyond the Scope is brought to you by the Digital Pathology Association.

You can visit www.digitalpathologyassociation.org for resources related to education, conferences, member benefits, and you know, all great things to get connected in the digital pathology world, which is of course the theme of this episode.

Special thanks to my co-host Giovanni Lujan and Abby Norris for their help in the planning and production of this podcast. And even more special thanks to the three presidents who were able to all get together at the same time across the entire globe, three different continents. This is awesome to have everybody together and record this.

We now present the panel discussion between June, Norman, and Esther, and it's filled with advice to vendors and pathologists looking to move the industry forward. Welcome everybody. Thank you for being with us today. I'm very proud of the panel we have today. Three very important people around the globe that are pioneering

not only digital and computational pathology, but also a new era in diagnosis in medicine. So I'm going to let him introduce himself. Let's start with Junya. Please go ahead. Okay. Hi, my name is Dr. Junya Fukuoka. You can call me Jun. It's like a name of a month.

I am a practicing pathologist. I'm doing general pathology and also lung pathology. You probably see my publication in the lung field quite a few. I am a professor of pathology informatics in Nagasaki University, which is located at the west edge of Japan. But also I am chairing the department of pathology in Kameda Medical Hospital, which is east edge of Japan. The distance between those two is more than 1,000 kilometers.

And I am currently serve as president of the Japanese Society of Digital Pathology, JSTP. If I say my current role in this task force is to bring the idea from Asian side and you know I this is a kind of Asia is so they have a wide variety you know

So, diversity, so I'm really thinking about how to disseminate digital resources. So that's kind of my role at this point. So I'm really looking forward to today's discussions with everybody. Thank you. Excellent. Thank you for accepting our invitation and for being here today. Norman.

Yes, my name is Norman Serber. I'm a computer scientist based in Berlin. I'm working at the Charité University Hospital. I'm head of research of digital pathology here at Charité, the Institute of Pathology. Besides this, I'm president of the European Society for Digital Integrative Pathology, SDIP.

and I'm chair of the IT working group of the German society. I'm also a long-term member of the standardization groups like IG Palm and Dicom Working Group 26. I think we will discuss this later on here as one of probably the interesting topics that we are facing for the future. And my role in the task force is mainly about

also holding up this flag of interoperability. It's, but also it's about of educating and disseminating knowledge towards pathologists with regard to IT and AI, but also vice versa to support the knowledge dissemination of pathology knowledge towards computer scientists, because it's very important that both groups understand each other. That's something we are doing in the framework of the European society, but also in the framework of some research projects that I'm

that I'm leading at the moment. And I think one of the last parts within the digital pathology task force, and that's probably true for all of us that I'm working on, is this thinking about what could be the holistic digital pathology workflow and how to come to this point that we bring labs towards using digital pathology. And there, of course, the workflow is one of the most important aspects to do this.

Thank you. Thank you for being here with us. Yeah, we're looking forward to a very interesting discussion. And last but not least, Esther.

Thank you, Giovanni, and it's great to be here. Thank you for the invitation and it's good to be here together with Jan and Norman. And we met a couple of months ago in person. That was really good where we discussed indeed the holistic approach, the global approach, how to move and drive the field of digital pathology.

My background is in biomedical health science, so I worked in the pharma industry, worked in pharma strategic product teams, proof of concepts, bridging R&Ds, and so forth. And why I'm here is actually, and not necessarily here in this podcast, but really why I'm here in digital pathology is...

My ambition and my goals is to contribute to making lives better, to improve quality of life, and also with that, that the patient comes first. And we all know that digital pathology, and this is in this era of digital health, actually, where we got more and more focus on digital, that that is extremely important that there is indeed interoperability.

And so what I did in the regulatory and standards task force, I served there for a couple of years.

But I transferred that to other persons who are now co-chairing the regulatory standards task force, Heather and Marcus, and are doing a great job. What we're trying to accomplish there is also defining general principles, how to get this interoperability achieved.

But also what my main focus right now is within the digital pathology associations, and that's what I'm focusing on right now as the president, is that we really are collaborating with all the stakeholders. So for me, it's very important. And also with that, for the entire digital pathology community, what we want to do is create this collaborative opportunities

I really believe in you have to move things together. You do it together as a team and not necessarily only within one association, but you need multiple associations. You need to do it globally. So that's why I'm very excited to work with allied societies. That is one of the goals, the main drivers.

as well as how to get digital pathology, how to move this forward and how to get markets access. And I believe together with Norman and also with Yun, we had our first real in-depth discussions a couple of months ago. And I really believe that we can have, if we continue this collaborative opportunities, we can make something work and we can kickstart this. And that's what I'm really looking forward to it.

Excellent. Thank you very much. Thank you to all of you. And you heard it. We have three presidents, three societies, three continents, three perspectives, very important perspective. So we're going to ask you to discuss something and then we're going to start passing the ball to the next panelist. And of course, if there are questions or anything that one of you have for us,

with respect to another one's intervention, feel free to ask them live. Now, I would like to start with a brief description of how you see digital pathology adoption in your neck of the woods. How do you think people are reacting to it? Are the speed of adoption, which are the most prominent use cases?

is primary diagnosis, research, et cetera, which one is the one that is heading or the one that you are more focused on or the one that has more challenges, just in general description of the use cases, the adoption in your respective areas. Let's start with Norman this time.

Yeah, I think that's a very, it's a very broad question. So there is, of course, historically, the application of digital pathology, especially that started in the educational area, right? So this was, besides research, this were the two most important areas.

because clinical always opens a whole new dimension of questions when you want to use digital pathology. So, of course, a lot of people in Europe are doing digital pathology in scenarios of education, but also of continuous education and also with regard to the

to research questions because there you have the possibility to investigate within large cohorts. You can do multicentric studies very easily. You can do even pan-national studies based on digital sociology. So this is really where a lot of initiatives and actions are taking part. There's numerous projects all over Europe, actually all over the world taking place. So I think we have already reached quite some

I would call it establishment of digital pathology. But I think where we are all aiming for is not only this, we are aiming for using it for the patient care in a more direct way. And this is when it comes to primary diagnosis or even before. So like what is also quite established already is like

second opinion and tumor boards. This were the first virtual microscopy applications that we had. I think that's also true for consultations between different countries, even between different continents. I know about some colleagues who were doing from Northern America to Africa or from Europe to Africa, but also towards Asia, there's connections. So we have a lot of this on national and pan-national, pan-continental levels.

What we see in the last months, maybe even in the last couple of years, is that more and more in the primary diagnosis, certain labs start applying digital physiology. We have the first solutions that are cleared from the American market, from the FDA, but also on the European market, from the European commissions, some

fields, which might be as an example for what we can see. And this is prostate, where we have quite some applications. And it's also breast, where we have quite some applications. Prostate, it's more about the questions of detection and grading, maybe.

and in breast it's more questions of quantification. So there's quite some applications, but once you want to start with these applications, you have to overcome all these hurdles, challenges that are on your way.

naming them like digital pathology workflow, interoperability, regulatory questions, reimbursement questions, and so on and so forth. And this is, I think, where we are at the moment in Europe, especially because that's the field I can oversee at the best. But I also know, and probably Abby and Junior will tell about

the Asian and American market as well, that we are really there only at the starting point because there are so many questions to answer and so many heads and minds to convince and to overcome hesitation that people might have. So I think this is where we are at the moment in Europe. Very interesting. And you're right, the question is kind of broad. So let me just ask you something

On top of that, what do you think is the main obstacle right now in Europe?

as a whole, as a continent for the adoption of the digital pathology in clinical use. And just because of the nature of Europe that is united in several ways, but also they have separated identities, is there a, do you see a big difference between certain countries and another in terms of those obstacles?

Well, so I mean, first of all, in Europe, even if you have from a regulatory perspective, you have the European Commission who is providing the boundaries for the application of such algorithms in Europe in general. So this is something that is the same for all countries.

But when it comes to questions of legal and reimbursement, you are dedicated to each country because every country has its legal terms here. Every country has different reimbursement mechanisms, is on a different level because of the payers that are here.

working on. And this also maybe answers your first questions, because this question of reimbursement and also of what do we need to do to really be able to use digital pathology in a secure and

proper regulatory and legal framework is something that has to be solved at first. Of course, there is, so this is, I think, one very important obstacle or hurdle that I see. But I see a second one, and that's actually, on the first, when you approach to this question at first, then you see that actually this should be even easier to overcome because it's only in the minds of people, and this is that people are having

hesitation that people are a little

feared about how this change might look because they know how to diagnose, they know how to do the laboratory workflow in the conventional way. And this is something that has established and evolved over the last decades. And there is a very good perception that everyone feels secure and is very comfortable with the way doing. And why to change things now? This is the question that a lot of people are asking. And

Having on the horizon, the image or the vision of that there might be computerized solutions and AI that helps you and supports your decisions is interesting for a few people, but it's for the most people, it's not enough to say, okay, I'm leaving my own

old habits and moving towards this new era. And so this is when you approach to this aspect even further, then you notice that establishing this digital pathology workflow is much more about convincing people and

taking people, bringing people together who have already experienced this next step, showing best practices and bringing those who you need to, you might need to convince together with those who have already experienced this new approaches and new possibilities because that's actually what this digital pathology workflow implementation actually is. It's not directly about

being quicker, being better. It's more like opening new chances for the future. It's about quality control, quality assurance, where you have a lot more possibilities than if you're doing this on a conventional level. And this is, I think, what the obstacles are and also what the benefits of overcoming these obstacles are. Great. Let's go to Esther now. What

So I'm going to sort of reframe the question into two parts. In North America, how's the adoption going in general? What are the main use cases? And then on top of that, what do you see as the main obstacles for the adoption of a clinical primary diagnosis workflow?

Yeah. And I think it's very interesting to hear as what Norman was saying, very much similar in the U.S.,

So let's start first with your first question. It's really about where is the adoption and what is the use case? And I believe, and I think when I speak with pathologists and what I see also from the DPA and in industry, is that we all know that digital digitization, that is the future, right? We all recognize that data is key. It's a key ingredient for health care.

So it is there with digitization, you make things more efficient. We all know the advantages.

The adoption is picking up. That's something that we do see as well. And the pandemic contributed to this. And deviations, for example, outliers and problems, they're always there and they actually needed to create opportunity. And that's what COVID actually did, right? It created an opportunity to have pathologists using it for remote sign-up.

and had to continue delivering the care to share knowledge while keeping mankind safe. That's something that COVID accelerated actually more or less. So we all see the benefits, we all see it in the adoption picked up, especially for primary diagnosis, remote sign out,

And also, in addition to that, as Norman was also mentioning, collaborations, sharing knowledge, helping each other, and reaching out to your fellow pathologists, to your colleagues, to expand your brain power, so to say. So I do think it's similar as what you see in Europe. It's similar in America.

is that it is picking up, the adoption is there, but we need more. There are also hurdles to overcome. And what you see happening is indeed the costs, even though within the US there's of course the reimbursements. Yeah, it is managed and governed in one way, as you would say, all the states in the United States.

are working with CPT codes, but that doesn't mean that that will solve the issue that you have for each state's specific regulations, rules. You have specific payers who have different procedures themselves and different processes. So there is a big challenge as well. So therefore, I believe that the adoption

within digital pathology is really also the costs. And costs are not only with regards to how much do we get reimbursed, but also with regards to how much does it cost to integrate within your entire healthcare system.

How do you make it fit in your own workflow? Because even though you might at a certain moment, you might have the regulatory adoption that that has been overcome, that that hurdle has overcome, that you get an approval and that you can use it according to the FDA, for example, then you have also the regulations within your clinical organization. Then you also have the integration. How can you

have the interoperability working? How do you integrate within your own workflow, making it efficient within your own lab? How to use it? How to ensure that your pathologists are using it and your histotects are using it? How do you make sure that your IT personnel is involved as well?

So it all comes down to help each other there, to make sure that people are being taught in a similar way, have a good understanding. And then it's more like, how do we implement it? So you have your capabilities, your assets, you know how to work with it because that's being now shared. But then it's also how to implement it, how do we really make it effective?

So yeah, coming back to your question, like what is then I think the main hurdle within the US?

And I think it's costs. I really think it's costs. And I think we can overcome it by having, we have the education, people start using it, showing it that how you can address it, that will also eventually reduce the cost because things will become more efficient. People get to know it. Yeah. And also you can come up with more evidence

to show that economic factors are not really a big hurdle either. So I think it's just that part as well, that we have to overcome the costs, education, getting people involved.

and learning from each other. That is how we can overcome it. And that's why I'm so excited as well. Yeah, and I keep reiterating that, but that's something why I'm so excited that we're here working on this globally, because I think we have the same problems and yeah, you can only solve them together. - Great perspective, Esther. Thank you so much. Now let's move to Asia. Junya, what do you think? Use cases.

Okay, yes. So adoption is actually just almost similar. I think even worse in Japan. The funny thing is we started quite early in the days of looking back, like let's say 2010.

Even before that, there was a kind of movement, small wave of the digital pathology installation inside Japan. The government is supporting the institution to purchase scanner. And at that time, there was more than 100 scanner already, you know, spread the

was sold inside Japan. But what happened was at that time, the quality of the scanning was not that really great. It takes about almost one hour to scan one slide.

So, you know, it's almost impossible to think about those make diagnosis and stuff. And the number of the slides, you know, are okay for the scan for a day is limited to five or six. So, you know, it's a completely different occasion compared to current status. And so since then, there was not much change actually.

Many of the reasons was, but yes, Japan is also saying it's a main issue is the money. How to get those reimbursement? That's very difficult, you know, other questions. And also on top of that, what is the motivation for the pathologist to install the digital pathology?

If they like something new kind of devices, you know, there are some sort of those people who likes those in a kind of advanced technology. But those people are limited, actually. Those people are actually limited and actually those are more focused on the university or big institutions. So then that the majority of the hospital, it is a middle size or small size hospital.

they're not much interested in about digital pathology because digital pathology do not change their workflow or any, do not give much benefit to them so far. There may be, but they don't notice it. It's not visible to them.

So I think that, yes, because we have to do the educational course or whatever to let people know what's the benefit, the true benefit of the consultation or true benefit of observing digital pathology, like let's say.

observation with low magnification, which is impossible by microscope, but which is beautiful by the digital pathology. And also some image analysis, like counting K67, HER2, whatever. So those are quite difficult by the microscope, but they're so easy by image analysis.

But then on top of that, I think that the necessity is probably a stronger demand and also a request from the clinicians. I think that's a very, or patient even, that's probably the important part which is now missing. So I'm wondering all the time, what is really making this realistic?

probably one easiest point is getting something, kind of an image analysis, AI model, whatever, digital pathology calculation as a type of companion diagnosis to some of the drugs, some of the medications. I think that those kind of candidates or models are already there.

something like counting PD-L1 or those type of marker is probably better to be done by the image analysis, not now, but in some very near future, I believe.

And maybe some of the medicine, some of the genetic, I'll say, the test or whatever can be only applicable to the patient who has enough amount of the cancer cell. But if patient, I think at this moment, nearly half of the patient cannot reach to that amount. So they just ended up to have a small amount of the analysis,

gene, I'm gonna say mutation tests or whatever. So I think if those can be supported by the digital pathology and if those, the software or models can be upgraded or recognized as a medical devices,

then things would be totally changed. It's gonna be a quite a strong demand from the clinicians and patients. So then we have to consider instillation of this pathology. I think that probably that's kind of my personal, maybe many people think in the same way though, but without that, the adoption, accelerating adoption is not possible.

in not only Japan, probably all the other countries may be difficult. At this point, it's the same scenario to Europe and America that, you know, what is the use case? Most frequent use case is education. It's more like an archiving purposes. There are some small amount of the people like me, like us, is doing a digital diagnosis. But I would say amount of the digital diagnosis in South Japan is probably less than 1%. It's a very, very small amount.

They use it for the conferences. Probably more people using the conference, maybe 10%. I don't know what the exact amount is, but the primary diagnosis is quite a high hurdle. So I think probably we need some sort of the ROI, which is the one I said was one of the examples. There may be several more, though. I think that those are quite necessary to really think about the adaption.

So let me just ask something there. Do you think that patients are influencing the practice of medicine in Japan to a certain degree?

I think that the voice of the patients are important. I think that, you know, like let's say for the cancer patients or cancer society, they usually have those sessions to the citizens, to the patients, or to the group of those people.

And that's a very strong voice. We pathologists have a little of those affections. Probably we have to do something in cooperation with a clinician or whatever to really have the voice for the patient. I don't know how to do it. I have a little amount of idea what is going to be efficient. But input from the patient as well as the clinician is probably important.

Let me ask the same thing to Norman and Esther. What do you think about that perspective in your respective areas? Norman.

Yeah, here we also have to understand that Europe consists of a lot of countries and this is very different in the single countries here. Because just to give one example, in Germany, for example, it's not very common that the pathologist has any patient contact at all.

So you have more or less the oncologist between the pathologist and the patient.

This means this direct interference between pathologists and patients is not happening that much. Of course, there is some cases where you also have this, but in most cases you don't. This is very different in Southern European and also in other European countries where, for example, the pathologist still is doing FNA, for example, and so has a direct contact with the patient. So this means

the influence is based on this different in all countries in Europe. Nevertheless, I would also like to add something to what Junya said before regarding the importance of going digital and what could be the benefits. Because I think it's much less about being more efficient and saving costs. It's much more about also approaching to

to this fact of we have the shortage in workforce, we have the growing complexity of diagnostic tools. We do not only have the histomorphological examinations with all the beautiful markers and more and more coming up and all those panels that we could use, but we also have the molecular questions here.

and all the molecular data and to incorporate all these data at the, let's call it the fingertips of the pathologist for a single clinical diagnosis for a patient. I think this is what is very, very challenging for the future. And this is something you can actually only do when you are working digitally

Because to incorporate all those data together, you need to put this in some kind of digital pathology cockpit and this requires digital data and it's not possible doing this on a microscope or in a conventional way. Well, it is probably possible, but it will take you much more time. It's much more laborious. And so I think this is another good argument why going digital is so important. Sure. Esther?

Yeah, and to add to that, what I also believe is that all that data is available. We know it's there with digitization. The question is for us also to discuss together, what are we going to do with it, right? What is it that we want to accomplish and how do we want to use it? I think it's there, but now we need to answer that question, like what is it that we want to accomplish? And the patient comes first.

So therefore I also think that a patient is becoming more and more important also in the US as you see globally. Patients are self-educating, a lot is available on the internet and they go with specific and directed questions to their healthcare providers.

Also, the pathologist is getting more and more involved in that. And you see it now and then happening, not that often yet, but I've seen some cases where patients really also ask about images and would like to be educated about what is it that their images contain. So that's something that I think that patients are becoming

becoming more and more important and driving healthcare. And not only from that perspective to want to know what's going on, but also with regards to

How am I being managed? How am I being treated to really understand that? And you see that now also happening in regulations. It's already there for a long time looking at patients reported outcomes, but it's getting more and more important. They're getting more advocacy groups. You see getting involved in discussions with with FDA, for example.

Also, with regards to real world data, collecting that real world data, patients need to understand what is happening with my data. Is it being used? How is it being used? And what does that mean in the way I'm being managed? And can it help in the way I'm being managed? So, yeah, I do think that patients are influencing more and more.

Of course, there's something that we also need to take account is the socioeconomic factors. And that's something that we need to work on as well. You see there are differences in different areas, rural areas, for example, are underserved. So now we can stand up and help those areas as well and help those patients also to educate them what's going on with their data and their medical history.

Very good, very good. That's a very interesting angle that as the usage of pathology for images, digital images spreads, I think that

a lot of questions will come up that will need to be answered from a regulatory perspective. And bringing the patient into the equation makes me wonder, what do you think those regulations will be in terms of patient ownership or rights to claim that data originating from their diagnosis, from their disease?

back to them. Like traditionally hospitals have kept the tissue, use it for teaching, for the slides, you know, make the rounds all over because they just, it's very easy to de-identify, but now you have all this metadata attached to the images. And what do you think will happen there? Will the hospitals, the departments, the pathology departments still be

responsible for that data exclusively like they are right now. They can do pretty much whatever they want with it. I think at least I believe in the US, I don't know if in Japan and in Europe, but the department is pretty much

the pathology departments are the gatekeepers of that issue and all the regulations that apply to that issue go through the department of pathology. Now being this more broad, that includes, it's basically out of the jurisdiction of pathology in terms of who can access it because now will be available. So who will be dictating these regulations?

do you think, I know this is down the line, but what do you think is going to happen there? And now with the patient on the equation especially, will he be able to say, I don't want my tissue to be anywhere or my images to be projected anywhere or to be used in any way? For example, let's start with, let's continue with estrosis. We were there with you.

Yeah, that's a very good and tough question. And I think a lot of ethics are involved there. So besides that, it has to be regulated, what's going to be done and how can data be used.

So I think what we will see happening is that that will evolve over time and just like with clinical studies, good clinical practices there, what can you use, how to anonymize data. And that's also the same in clinical practice, so not only in the clinical studies.

So yeah, it is a real tough question to answer. I can only speculate. I have no idea where we're going, but I do believe that the patient of course is the ultimate person who can say what happens with their data and how much data will be disclosed. And it's up then to the vendors, the industry,

but also to the healthcare providers to make sure that that is being, yeah, that that is being secured. So that's, that's my personal opinion. So they should be careful what to do with data and how to disclose it to third parties if, if disclosing anything. And yeah,

Yeah, so it's a tough question and it's an ethical discussion. But I do believe that how it's set up at the moment can work in the future as well, that you always separate data. You always ask the patient for consent. And what I mean with separate data is that you always make sure that the metadata is not shown to anyone else without the patient's consent or that you always anonymize it.

So I think those principles are there and those are good principles, in my opinion. Absolutely. Juni, any opinion on that? Yeah, this is a very, very difficult question. Tough question. Yes, you know, if you think about this, if patients hold those digital data, what would happen?

Certainly, it's so easy to consult to whatever the prices. And if the AI is available to the patient, they may use those AI to evaluate the tissue, whatever that is. So I don't know if that's kind of happening or not, but my kind of worry is, of course, there are some increase of those legal cases, which I don't want to be the happening, but

But there may be such kind of worry would arise, of course. There should be some regulation on that. And I think there should be some kind of discussion how to deal with those is important. And for especially, let's go back to those pathologists who is using a microscope. But if patients hold the digital data, if they demand, they can do it, right? Then what happens?

So that in a microscope and observation, I would say it's quite kind of dangerous, right? Because so microscopy and observation, they cannot mark what the evidences are. And if the digital materials are copied and spread into some of the other doctors,

Well, that's a kind of a very complex situation. So yeah, I think that this, I didn't think so much, but now I just got those idea. We probably have to have, seriously have to think about those regulations. Well, I don't know, just much ideas, but this is a kind of idea popping up in my mind now. Yeah, just something to think about it as the patient gains that access

And then they have images now in their computer that they can probably analyze with, I don't know, Google might come with something that will do some analysis and they can start trying and that might be confusing for the patient. So we know that sometimes limited knowledge is very, or this peripheral knowledge can be very dangerous.

So yeah, a lot to think and do about that down the line. Nor about anything that you want to add to that.

Yeah, I can definitely underline that this is a very, very complex question. So we could probably spend the whole podcast only talking about this question because it has so many facets, starting from the GDPR question or the question of how to secure my own data and how and why.

and which ways I can make them available and to whom, and how can I make sure that I can always have the control over my data. Going to the question of how can consent to the hospitals be made, there's already mechanisms of broad consent for some data or of a very narrow consent to the data.

to see what exactly can be done with my data. There's also the possibility of, or the question of all the ethical committees that are required for all studies that are carried out to be accepted before the studies are even starting, and to access all those data and specimen that are available in the hospitals. So there's a lot of in these aspects, but I also think it's very important to,

to see also the difference between academic research and industry driven research. Because when we are talking about, or like you just said, the pathology departments or the pathologists have all the data. And then in the next sentence, we are talking about AI being developed. And that's actually two completely different worlds, right? Because for when we are talking about clinical decision grade,

AI algorithms, then we are talking about companies who are developing this. I'm not talking about lab-developed tests where one institution has a few computer scientists working together with a few pathologists developing some algorithm that they use in their department. I'm talking about those who really are used in several departments. Therefore, you need big clinical studies, you need a proper quality management, and so on and so forth. So this is something that the

that a company has to do. It's nothing that a research group can do. But if we are talking about fundamental research, this is something, this is some academic research. And that's a complete other story of finding new patterns, of investigating certain diseases.

learning more about certain diseases and we should really distinguish between these two worlds. And this is already what is done in the past or what has been done in the past. And we should also do this in the future. So I think there are quite some mechanisms

in Germany, but also in Europe in general are in place. I think it's not very different also with how data are going to be shared in the United States and in Japan. So I think we have some good starting points that we should make sure, of course, to give the patient the control over their data. But I mean, the scenario that you just sketched, like the patient has his whole slide images and can upload it to some database

to some whatever application, get some quantification and do his own diagnosis without any background that's more that's that's for me like like a sketch from a horror movie yeah because this this can only lead to to wrong information to wrong um

And then when the person, when the patient has learned about my own investigation, this, I'm having this disease, I'm in this grade. And then he goes to his doctor, to his pathologist or to his oncologist and says, well, look, I have already learned this endo-unconsciousness.

said, no, no, it's very different because you haven't taken this and this into account, then it's more like convincing the pathologist that the patient, that the pathologist knows better. And this is something we don't want to get into because that's the very wrong way. So we should be very careful. Of course, patient empowerment is very important and give the patient the authority also over their data, but

Let's be careful that this is not turning against the pathologist and the pathology as a domain at the end of the day. Yeah, because in the end, we'll be against the patient themselves and their well-being. Yeah, that's interesting. Now, this is a good point to say way into standards. Standards is, let's first kind of,

I would like to hear your definition, your opinion on the standards and how are you doing in your respective societies? Are these standards expected to be global since we know most vendors kind of produce their products to be distributed globally? Or do you think that there will be more and more

self-regulated by each country, by each society. Let's talk about in general standards, what are we doing, what's the more important and what do you see happening in terms of standards? That will be sort of the guideline of the discussion on this topic. Juniel, let's start with you.

Yeah, standards is actually different by the organs and disease. Some of them are more like a university, you know, the same. So you have to follow the, you know, completely the standard, like let's say, WHO classification to classify the disease. And I think that's how most of the, you know, for most of the tumors and diseases. But

At the same time, there are several conditions which is different by the races. For instance, let's talk about lung cancer, then the mutation status is completely different.

Right? We Japanese, half of the adenocarcinoma has EGFR mutation. That's not the case in the American Europe. So it's a completely different type of molecule or tests are required for the patient's populations. So I think that the standard has to be also evaluated in their own country is important for some diseases.

especially for something related to the medications. But generally, the diagnosis has to be standardized. And so that then a quality of the scanner and quality of the specimen, many things also has to be standard as well. We probably need those kind of a statement how to standardize throughout the world if we think about this.

Yeah, I'd like to add here that this is also something very special for the pathology domain because

Passology comes from optical light microscopes. It comes from stainings being done with, of course, by SOPs in a lab environment. And now we are coming to a point where we are applying digital techniques. So they are very deterministic. They always do it the same way. The algorithms perform, I mean, despite the aspects of some AI being

being a black box, but it's very standardized, right? And you are trying to apply something super standardized like scanning device or an image analyzers or pixel based image analyzers to something that has been done manually and that is a little different. So like if you are taking a look, one concrete example, if you're taking a look at the agent E-stain

of different laboratories. And you have the exact same specimen being stained and cut and stained in different laboratories. The outcome is so, so different. And that's actually something that now in the second stage,

of the digitization some algorithms should deal with. And this is like a chicken-egg problem. What do you want to change? Do you want to change the algorithm? So do you say like the algorithm has to be capable of these variations? Or do you say, well, we should probably standardize the H and E stain so that the different labs have the same outcome for this stain. And then the algorithms will have, of course, an easier game.

But adding to this is also that this also does not only apply for the algorithms, but also if a pathologist is moving from one department to another, they really need some time to incorporate themselves and to understand how this new stain is working. And really talking about the H&E only, which is the super blockbuster in pathology, as you all know, there are so many differences with this stain. It's

It's incredible to see that there has no standardization being carried out yet. And so this is also a question that we should discuss. Do we need maybe standardized things? And should we do something here also on a national, continental, or even global level to do this? Because when we are thinking this whole story of digital pathology end-to-end, then we should also discuss this. And this is only...

only and probably Esther will now tell about also some technical parts because this was also what I had in mind when you raised the question of standards. Of course, Junior started as a pathologist with all the medical domain, but of course there is a whole new level of standards also in the question of interoperability and this also requires standards, but maybe Esther, you can take it on from here because you're also very much involved in that.

Yeah, and I think you're absolutely right also with standardization in coding language, for example, what we use is inputs to develop our algorithms. So if you think about reported outcomes or confirming what type of diagnosis has been used, are we using the same terminology, the same diagnostic codes for the same diseases?

globally. So that's something that we also have from that perspective in developing a medical device, we also need to understand that how do you call a feature, what is the code for that feature, how to annotate that particular image so that your inputs into your algorithm to train it will be the same.

Of course, within looking at scanners from that technology point of view or also for algorithms, you can use standards that you use for color. So I think there we should use whatever is already available, standardize as much as we can, use the global and the general principles. So that makes it then easier to develop your in-vitro diagnostic products from that perspective.

So DICOM, we can start building that towards more to an accepted standards. It will not solve the entire problem, but it's a small step. We do the same thing with regards to other standards for assessing risks.

We're all using the standards to how to assess risk for a medical device. And then that will be used to develop your device. So we can all use those standards that will make some testing that you do to develop your device more according to general principles and be understood.

That those are the first steps that we have to take and then we can take it from there. But I yeah, I think it's great. And I also think it's great that we started with you because looking at from the pathologist perspective.

How do you standardize there? Do you want to standardize because everybody's using a different workflow in their own hospital as well? Is it really that we want to standardize that globally? That's one is like, of course, a big undertaking, but also, is it really needed globally?

So from that perspective, I also think we have to look at what can we standardize, what should we standardize? And then it's more like, how do we do it so that the end user and also ultimately the patient will benefit from it? I think that is the question that we have to ask and then define what standards should we use. That's a great summary. Thank you all of you for all those questions.

those perspectives. I'm going to pick up something you said there. You mentioned iCOM and that's the image standard that radiology has been using traditionally for many years. I know that they had their own struggle at some point, but finally they decided to start using one standard for their images. And it's been standardized for many years and they are

I think it has worked for radiology. Do you think pathology will have the same standard image format, either a diacom or something else? And do you see this as a global effort? Because obviously, as we mentioned before, the vendors, the ones that are making the equipment, are marketing this in the

in all the continents, they might be Europeans or American or Japanese or some other place of Asia, but they market them globally. So let's talk about that particular stand, image format. Norman, what do you wish to do?

It's a well, when I first joined, I think my first DICOM meeting that I attended and when I joined the working group 26, who was particularly dealing with pathology images in the DICOM field, it was in 2012 or 11 or something. So and back then there was already this feeling in the group that

"Yeah, now we are there. We have this first standard and now we can start using DICOM." And there is the scanners out there. They can of course produce DICOM and yeah, let's go. Now we have a standard and everyone can use it and it's interoperable.

Now we are like a decade later, more than a decade later, and we are still discussing with vendors, yeah, you're creating your proprietary formats and when incorporating them into APLIS system or image management system, then it's still like talking about VNAs and not DICOMPAC systems. So I think the

the way to go towards DICOM is the most important and the ultimate goal that we should all have. But the problem that I'm constantly hearing from within the group is that

It has to be the users at the end of the day who are requesting this from the vendors because the vendors will not do this, will not go towards DICOM because it's an additional effort. They have their own formats. They started coming from those one-stop shops where they can incorporate solutions from their own portfolio to

together with another product from them. So this is where these companies started from. And this does not require standardized formats. But now we are already at a point where we see that the number of solutions and the number of combinations of solutions is so high that it requires some standards. And here, it's not only DICOM. It's also this what the IG Palm working group is doing because it's much more about the workflow.

it's not only about the scanner is producing an image and then it's stored somewhere and that's it. No, it's about how all the metadata is incorporated to those images, how we can have the information about

what is the specific profile that was used, why it was digitized, because that is at the end of the day important when the AI or the image analysis algorithms are running on those data. They also need to incorporate all of those metadata and all the clinical data also has to be incorporated. So it's a whole big thing of data that

that needs to stick together and this is where then the standards and those standardization comes into the game and becomes more and more important. So to summarize, I think DICOM is definitely the way to go and the working group is growing

We are doing a lot of additional things. Also now we're having annotations and computerized annotations being in the standard since mid of last year. So I think it's the way to go, but we really need the question and the demands from the users, from the pathologists, from the decision makers in the hospitals, in the clinics to ask for this interoperability. Otherwise, we will still probably take another, I don't know how many years until we are there. Excellent.

Esther, what do you think about that? Yeah, I'm always wondering, is DICOM really the answer? And I'm always promoting it, don't get me wrong. But I am wondering, is it the answer? And Norman touched upon very relevant things. It's not only DICOM, there's much more to it. How do we integrate data? How do we use data?

Looking at it from a user perspective, I do think they are asking for it because what users are asking is actually, we want to have an open system.

And an open system doesn't mean to them an open system where we can have a system from vendor A getting all different modules and then having algorithms attached to it. I think what a user is asking, can we have

a scanner from vendor A, an algorithm from vendor B, and maybe an algorithm from vendor C as well, and then maybe working it with an LIS system or our own hospital, Bruun, own lab information system.

So it really comes, in my mind, it really comes through digitization, how to implement it and how to use it in your setting. So I believe we have that question from users that they do want to have open systems. They want to have interoperable systems, so to say, instead of those closed applications.

Looking at it from a vendor perspective, I do believe we have to work on this and we have to answer those questions. We have to ensure that we can work with all this. So standardization general principles are needed. Also from the FDA, I know that they're working on it. They have this entire digital health center of excellence set up where they are looking into how can we

apply all this digitization that is happening not only in pathology but everywhere.

So, yeah, so that's coming back to like do we really need DICOM? I think yes, so I'm promoting it. But is that the only answer? No, we need much more standardization. And there again, we need to listen to what is it, what the user really needs, so that it will be possible to put it in their workflow.

And with that, we straightaway will work on adoption because I think we're creating our own barriers by not working together and also not seeing that standardization will solve these problems. When you use standardization, everybody can use whatever they want, what kind of elements they want, whatever modules, and then the adoption will be bigger. We will accelerate it. Thank you. Thank you for that.

insight. Junya, anything? Yeah, I don't have much actually. I am a pathologist, so I don't actually know much about those technologies, but my hope is, of course,

standardization standardize in format as necessary. So for that, I think DICOM is probably important. In terms of the storage wise, if pathology is in a separate storage, it's probably create some problem in the future because storage is quite large. So it's better to be, you know, melted into the whole, you know, the medical, you know, the storage

along with the other radiology. So in that way also we better use those packs and DICOM. But yeah, so far there's not much viewers or applications that enable us to observe the DICOM. So there are some vendors who can create the DICOM, but there is not much usage.

So I think that the direction is of course to the applications, more user interface, more kind of the appealing applications which are only available to the DICOM is probably necessary as from the part of the view of the users. So not much else. Thank you. Thank you so much. Thank you to

Maybe just to add one thing because this question also makes it was a very good summarizing question for this collaboration of the three societies. Sure. Because what SDIP and DPA and also in the future the JSTP are already doing is supporting all those initiatives from the DICOM Working Group and

from those interoperability groups, like having the Connectathon tour carried out at the past visions and at the ECDP for quite some years now. There's an upcoming Connectathon now at the past visions in October, and we are also trying to bring the Connectathon to Japan. So this is really something where these three societies are already working together in this framework of standardization interoperability. So this shows that

especially such a collaboration of allied societies between the three of us is really giving a benefit also to the standardization efforts. And that's very, very important, I think, because it's one of the most important topics to solve besides those that we have already discussed before. - Yes, absolutely. Esther, go ahead, do you wanna? - No, I just want to say absolutely, thanks for pointing this out. I wholeheartedly agree.

You know, this has been a great, great conversation and I'm thinking of so many questions to ask to each of you because it's so important to have a perspective from different angles, from different societies and from different places on this globe, how things are going, how things are doing, what has worked, what hasn't. So unfortunately, we're

running out of time, but we certainly would love to do this again. David and I will be live at Pathology Visions next month, and so we are hoping to maybe get an opportunity to chat with you individually or maybe with the three of you sometime soon. Last questions that I want to pose to the panel.

It's about a message that you will have for one, for the pathologists, for practicing pathologists, whichever the area of practice is, and for vendors. What would be your message to them from you, from your society to those in whichever part of the world they are? Let's start with Esther now.

Yeah, so for pathologists, I can only give my opinion coming also from an academic background. But sharing your findings, sharing your knowledge, sharing your research, that's going to help someone taking to the next step. So I think that that is, yeah, that's something that we can do there.

From Venra's perspective, yeah, and again, I'm a strong believer in sharing knowledge and

So what I would say is when you share knowledge, you will team up and come up with general principles, for example, what we just have been discussing. But also what I would like to say, and I'll keep it short, I would say don't be afraid of your competitor and thinking that if I don't share my knowledge, I keep it to myself as power. I am a strong believer that if you share knowledge, it's even more powerful.

That's how I would like to conclude. Thank you, thank you Esther. Great conclusion. Jonje? Yeah, it's the same actually. Aprizio noticed the importance of digitization

And because this is the future, it's not even future, it's now actually, but the future, digital pathology is very important. And if you get some benefit from the digital pathology, please share the knowledge, not only the knowledge, if you build up something, please be a reference site, become a reference site to invite people to say, hey, that's what we have.

You know, those knowledge is very important because people are so afraid because how much effort they need, they feel like, well, it's impossible for us to start.

It's quite a effort, but some people know about the tips by the experiences. Hey, you can do this more easier. We can eliminate these type of things. So those kind of tips are here and there. So if you are going to start up something, digital pathology or AI, whatever, just share those knowledge and invite people to show this

that's a kind of a message to the pathologist. For the vendor, do more hands-on. You know, people need to touch. So, you know, more hands-on to the vendors. That's my message. Thank you. Thank you very much. And Norman, what's your message?

Yeah, I think the most important things have been said. So I would maybe towards the pathologists, I would give the question or the request of do good and talk about it. So those who are doing digital pathology, share your knowledge, share your experiences so that others can learn from you. Also that others don't make

mistakes that you might have done or that went into some dead ends that you experienced. So we should really learn from each other. And this can be only done by sharing knowledge. Also with regard to the reimbursement, which I think is a very important thing. I think this also has to come from the pathologists. They have to team up with the patient advocacy groups and with the professional associations to approach to the payers and really make

reimbursement possible for digital physiology in clinical use for AI application in clinical use, because this is also a catalyst for bringing AI into the clinical diagnostics.

And towards the vendors, it's all about interoperability. We have to overcome the situation of implementing interfaces towards only configuring interfaces. And this requires that the vendors have to make their homework, that they have to

facilitate an approach to these standardization groups like IG Palm and Dicom and implement all the standards that are there already because then we really come to the point where we can stick together solutions and configure them and the solution and the whole end-to-end solution is available within a few days and not within months how it is sometimes at the moment. Great messages, you guys. Thank you.

Thank you.