How Are Pathologists in Japan Using AI and Digital Tools? With Dr. Robert Osamura

Aleks: [00:00:00] How did your colleagues like the transition to digital pathology?

Robert: It varies according to the individual. Somebody still sticks to the microscope. I know we need some time to, everybody’s getting used to implementing it. I didn’t feel much difficulty to get the machine and start looking at digital images.

And as far as the reporting is the same format, we have to look very carefully the digital image. We have to be very careful that we really watching the exact same changes in digital image. We can feel more secure. I mean, the AI doesn’t replace pathologists, but we feel more confident that secure.

Aleks: I think you guys were super motivated if you still have to do glass, but also did digital pathology.

Robert: Yeah. Yeah. Right.

Aleks: So that is. I think an important step when you wanna drive adoption of something and also gather data to show the regulators [00:01:00] that hey, at some point we can drop the glass.

Robert: I think that the future of pathology is more patient-oriented…

Aleks: This episode is one of our special USCAP episodes that were all sponsored by Muse Microscopy, so let’s dive into it.

Aleks: Welcome my digital pathology trailblazers. I have another guest for this special edition podcast at USCAP. This is Dr. Robert Osamura from Japan. Welcome Bob to the podcast. How are you today?

Robert: Okay, thank you.

Robert: My name is Bob Osamura and I’m a practicing pathologist. At the community hospital now, and I have worked for IAP International Academy of pathology and International Academy of Cytology.

As well as the Digital Pathology Association, DPA. And I’m really interested in endocrine pathology, breast pathology, and digital pathology. [00:02:00]

And particularly recent application for the AI artificial intelligence is, very attractive and we should think of how we use efficiently in our future practice.

Aleks: So currently what’s the place that you practice right now?

Robert:  I’m in, at Kokan Hospital, Kokan Hospital near Tokyo. In Kawasaki. Yeah. Uhhuh. Yeah.

Aleks: How big is your hospital?

Robert: About 300 beds. Okay.

Aleks: And is it a big one in Japan or compared to others? Is it a big institution?

Robert: It’s a small, rather small size hospital. And we do a practice in pathology in various tissues in kinds of tissue. And we apply digital pathology in small biopsies.

And we read them every day. Yeah. [00:03:00]

Aleks: So you use digital pathology every day for your normal pathology job, right? For biopsies?

Robert: We yeah. We are the colleagues read the double check with the micro… microscope.

Aleks: They do. So you have a double workflow? Yeah.

Robert: Yeah.

Aleks: When did you start doing that?

Robert: About, 3, 4, 4 years ago. Yeah.

Aleks: And what is your workflow with digital pathology like, what scanners do you have? Just tell me basically how you implemented digital pathology in your institution.

Robert: Yeah, just regular work. Yeah. Yeah. Routine diagnosis. Small biopsy first.

And takes only short period to scan the image. If we work on the larger specimen takes more longer time. So we do only bio small biopsies.

Aleks: Okay. Yeah. [00:04:00] Yeah. And how many scanners do you have in there?

Robert: Only one. 1 

Aleks: 1. Yeah.

Robert: Hamamatsu 201

Aleks: Japanese scanner in Japan. 

Robert: That, it’s a very good, yeah. Yeah. We selectively, we scan cytology, but not, it’s not the routine work. Yeah.

Aleks: So when, why did you decide to implement digital pathology in your institution? What was the driving force? Why did you decide that you need to have that technology in addition to glass slides?

Robert: First I was interested in the digital image, how practically useful. So we started testing that and we realized that’s very good. To easy to scan. The lower power and easy to go to high, higher power. And I think it’s very efficient and still very enjoying [00:05:00] the, looking at the image.

And also you can send image in the hospital, like a conference and also an outpatient clinic, they can share image. Which cannot be done in a microscope.

Aleks: Yes. Or it takes very long time. 

Robert: Yeah. Yeah.  Also consultation. Yeah. If difficult case we can send out of course we have to be careful about the privacies and also all that protecting those privacies of patients, you know it’s very convenient.

Aleks: And so you guys have in Japan in general, a unique use case because you are a country of islands. So this connectivity that digital pathology provides gives you an advantage that is like more of an advantage than in a country that is basically connected by land [00:06:00] and have land-based transportation.

Robert: Yeah, that’s true. That’s true. So by digital image, you can easily cross the ocean.

So we enjoy that. Yeah. And also going back to the hospital, inside the clinicians can appreciate the digital image.

Aleks: Oh, so you do have the pathologist collaborate with the clinician and look at the images.

Robert: The Chart. Clinical chart is digitized. On the screen, in the computer. So you. The clinicians can click on the image, and they can confirm what we say in our diagnosis. 

Aleks: Yeah. Oh, amazing. You mentioned that your pathologist do digital pathology of the biopsies and glass in parallel.

Robert: Yeah. 

Aleks: That’s for the same tissues or for different tissues, [00:07:00] like you do…

Robert: Same tissue. Yeah. 

Aleks: So you have both a digital image and a glass. A slide for them to review. 

Robert: Yeah. There’s a regulation by the government. Still right now it’s getting more practically the use of digital image.

But up until recently we have to compare both digital and microscope images. 

Aleks: While you are evaluating one case, like in the same workflow.

Robert: Yeah. But there was a thought that, we have to confirm by looking at the microscope.

Aleks: Ah, okay.

Robert: But now it’s changing. Now

Aleks: Uhhuh,

Robert: So only a digital image can be.

Aleks: I think so. Because obviously this is different than in the US. 

Robert: Yeah. 

Aleks: Now you can do primary diagnosis without the reconfirmation on glass. I think you guys were super motivated if you still have to do glass, but also the digital pathology. Digital, yeah. Yeah.

So that is I think, an important step when you wanna drive adoption of something. [00:08:00]

And also gather data to show the regulators that, hey, at some point we can drop the glass. It’s good enough. 

Robert: That’s right. Yeah. 

Aleks: And let us just do digital. 

Robert: I think it’s about time too. We can look at only digital images. Yeah.

Aleks: What’s the process in Japan like? You would have to submit some studies or like how do you have the conversation with your regulators?

Robert: Yeah… There’s like a. FDA.

We have PMDA, government, which regulates those all rules or the diagnostic regulation and also reimbursement.

So right now, if the hospital does not have pathologist, they can scan the image and send the image to a larger hospital with a pathologist

So make a diagnosis, then two hospitals can share [00:09:00] the reimbursement.

Aleks: Oh, okay. Yeah.

Robert: Yeah. So that’s probably big advance, future step for the digital pathology in Japan.

Aleks: Was it…

Robert: But still long way to go. Yeah,

Aleks: I, in the US, it’s still long way to go, I think across the globe, and we’re gonna talk about it in a second because.

We are at the Muse Microscopy booth and they are the sponsor of this whole series of podcasts, and they’re like the next innovation. But before we go there, I wanna ask you, when you were implementing digital pathology, when you started your revolution at your institution, what were the challenges?

What was difficult, or did everybody like come on board immediately and like it? Or what pushback did you get? How did you go about implementing it?

Robert: Implementing it? I didn’t feel much difficulty getting the machine and starting to look at digital images [00:10:00],] and as far as the reporting is the is the same format. Yeah. 

Aleks: Okay. 

Robert: Yeah. Then we have to look very carefully at the digital image. First, we have to be very careful about. We really watching the exact same changes

In digital image. Sometimes very small organisms were difficult to see and we have to confirm by the microscope sometimes, but now we realize some of them really, when we get used to it, we, we can look at it, but still, I have a microscope always. So anytime. I have a little question, I can easily… 

Aleks: go to, you can go to the microscope. 

Robert: I think that’s very important.

Aleks: And so…

Robert:  Throw the microscope away, somebody said that… [00:11:00] 

Aleks: No. And when I was hearing at presentations or from other podcast guests about their implementation, like their journey at their institutions, they were, there was often a period for the pathologist. Even now in the US, you don’t need the glass. The pathologist can always ask for the glass if they need it.

And also, when they were starting the whole process, they would let the pathologist always have the glass if they wanted to. Yeah. And they the timeframe when the last pathologist would say, okay, I don’t need the glass anymore, would be six months. Okay.

Robert: Oh. What do you mean by that?

Aleks: So you would have a period of six months where everybody gets the image and the glass.

And the pathologist says they themselves would say when they’re confident. Yeah. Yeah. And drop out the glass. And the latest adopters would last six months with both.

Robert: Oh, probably. Yeah.

Aleks: And then they said, okay, digital. [00:12:00] But always when there is like a, something that they want to confirm, something that they need to check.

They can ask for the glass. 

Robert: Glass. Yeah. Yeah, that’s exactly what I’m saying, we have to, we can’t throw the microscope away. Yeah.

Aleks: I mean the microscope…

Robert: I don’t know, in 10 years,

Aleks: Maybe when we implement the MUSE, the microscope is also, so the whole slide scanner is actually a microscope.

It has objectives, it has all the components of a microscope. But creates a digital image instead of that, the analog that we see through the microscope. Uhhuh, how did your colleagues like the transition to digital pathology? Did you have any vibe? 

Robert: It varies according to the individual. Yeah. Yeah.

Somebody still stick to the microscope. Yeah. That’s okay. Yeah. I know we need some time to everybody’s get used to this stuff. But eventually I think, you know [00:13:00] we, we, used to teach the images.

Aleks: So now let’s talk about the MUSE technology, which is the direct-to-digital imaging from tissue sections. And where do you think that would help you in your current

Robert: In MUSE?

Aleks: Yes. If you would have the MUSE system. If this was like already to go what would you, what would be your application? Where would you use it first? Where would you use it second? Like what are the advantages that you see for your particular use case in your country, in your hospital, how you work?

Robert: Let me think. If I have a, that machine Probably I use it for the first, I apply for the intraoperative diagnosis.

And if the surgeons are waiting in the operation room and we. We can just put the tissue, fragment tissue and cut the plane. [00:14:00] Get the digital image.

And I heard there is a virtual color, it can be used.

Aleks: Yeah.

Robert: So I hope it, it takes a much shorter, so then we can reply to the surgeon very quickly.

And then the surgeons can go to the next step. So I think the first application I can think of. With inoperative. Rapid diagnosis.

But then if we can use a fixed tissue right, then we can use for the permanent diagnosis. So probably we can think of two ways, rapid diagnosis and also permanent diagnosis. Uhhuh, yeah. But permanent. Then right now we have a section stain. It’s a permanent, we can save it permanently, but if it’s a chunk of a tissue

Even if fixed, [00:15:00] we can keep it forever. 

Aleks: That’s true.

Robert: So then we have to think about it. But anyway, we can make a permanent image to keep at a digital level. I think it’s good. And so I know there recently there’s a virtual color. And even it’s immunohistochemistry can be virtually colored.

So everything, once you have a plane cut the plane of the tissue, we can start thinking about H&E. And based on H&E, we can make a virtual color, the various things. Everything is, is a virtual that, that’s probably one way we can think of. I’m not sure how much,

Aleks: We are hypothesizing right now. Yeah. Right now. But that’s the question I have. Yeah. I wanna see like, where do different key opinion leaders from different corners of the world, see this being applied and frozen sections [00:16:00] intraoperative is something that comes back over and over. 

Robert: Yeah. Yeah.

Aleks: So I wanted to ask you…

Robert: The… 

Aleks: Go ahead. 

Robert: Excuse me. The frozen section is, as it’s histology morphology, that’s a satisfactory,

Aleks: Yes, everybody said that they don’t like frozen sections compared to H&E. 

Robert: So I think looking at the MUSE technology, I think it looks better than the frozen section.

Aleks: Definitely. I think so too.

Aleks: And what I wanted to ask you about is about what you already mentioned, that some hospitals don’t have pathologists and then they can collaborate with hospitals that have pathologists. Like how many, like outside of big cities in Japan, how many remote areas do you have? Is it, is there a lot of hospitals in the remote areas or care places?

Is this a problem there, or

Or a reality rather than a problem?

Robert: Theoretically it’s a very good [00:17:00] system, but reality is…

Aleks: You guys are famous for good hospital in Japan. 

Robert: Without pathologists do have a do. I have to have the scanner

Aleks: Yes.

Robert: To, to the scan image and send it to the pathologist.

Aleks: So they also… 

Robert: So that’s it’s, sorry. It’s a very expensive scanner. So if those abuse scanner is less expensive,

Aleks: Then it’s a solution. Right?

Robert: Yeah. And also technologists can manage To cut the plane and send the image.  Then probably become very popular.

Aleks: I can imagine that as well. Also, you say that if they can send a slide to the pathologist, that means they need to have a histology laboratory, right? No. Or can they send tissues?

But if they send tissues, then they probably cannot share the reimbursement. Yeah. ’cause then everything goes to the hospital where you send it. Interesting. [00:18:00] 

Robert: Not scarce, secure

Aleks: If you have it, and if you can confirm it. Yeah. And I think several, many pathologists, especially the more experienced one, the older generation, they want that backup. This is what they trained on. This is what we all trained of. I’m a veterinary pathologist by training.

Oh, you? Yes. Now working in drug development. But this is we trained on glass. And many feel a lot more secure. Doing work on glass then On digital.

Robert: Yeah. Yeah.

Aleks: Even though the image is generated in the same way, it’s still light microscopy. 

Robert: Yeah. Yeah. And these image is good enough to make a this ordinary changes of morphology.

But some detail, very minute changes or organisms then, and I need some confirmation. Yeah.

Aleks: Yeah that’s [00:19:00] totally justified. And…

Robert: Then if you use AI, then if AI works much better, then we can just run the AI first, then let them pick up some dose small changes and then we can feel more secure. Yeah, AI doesn’t replace pathologist, but we feel more confident that security…

Aleks:  You have like a tool that already points out things, prescreens things for you. That’s right. And con… you can confirm or if you disagree, you can just disregard this and your own analysis.

One more thing before we wrap up that I wanted to ask you. What’s the, do you know what’s the percentage? How many hospitals in Japan are. Doing digital pathology, is it a lot? It not so many. 

Robert: It’s very good and difficult question to answer.

Aleks: And I know that in Asia, [00:20:00] it varies per country. A lot…

Robert:  Varies and still because of the very limitation cost of these scanners. And several other things. I would say still very limited. Hospitals use. A fully digital system. Yeah.

They might have a scanner, but… 

Aleks: Okay. 

Robert: How practically they use? I’m not sure, but it’s very limited still. So how we can we can go further. Probably we should be more familiar with the system.

And also scanners should be more popular. And can get in the lab. So maybe still it’s a very expensive right now in Japan, government approves [00:21:00] Hammatsu, Phillips, and Leica.

Aleks: Okay.

Robert: So they’re good price. Yeah.

Aleks: Yes. And obviously MUSE is a new device. Yeah. So that’s gonna be another regulatory.

Pathways differ in different countries. I’m excited about this. Let’s see,

Robert: But I’m looking forward to the using the digital images as routine work. And also if it’s combined with AI, maybe we can predict some molecular changes or more, give us applicable. Information about the treatment. Yeah. Yes. So I think the, in future of pathologist is more patient-oriented and more clinical information.

Aleks: That is also a theme that comes back again because [00:22:00] it’s a little bit, people are, pathologists are looking forward to having systems and devices and software. Make them more efficient. But not so that they can do an even more job, but rather so that they can be more involved in patient care.

Robert: Yeah. That’s, that that’s the point…

Aleks:  In interacting with clinicians and being part of the patient care team. Yeah. And not just this specialty that delivers a report and somebody else takes it.

Robert: Yeah. That’s right. Yeah.

Aleks: So I think also for our profession, this is…

Robert:  Yeah, that’s exactly what I’m thinking. And also many people are thinking, pathologists in the future. Yeah. Otherwise, no, no reason to digitize or, yeah. What the point. Yeah.

Aleks: I think the same.

Robert: Yeah.

Aleks: Thank you so much for joining me.

What I wanted to ask you yet before we go is where can people, where can the listeners learn [00::2[3:00] more about your work? You mentioned you’re active in Digital Pathology Association as well?

Robert: I’m a member of the education committee and… 

Aleks: Okay. Yeah. 

Robert: I joined the committee meeting and I’m available… 

Aleks: But okay. Yeah If anybody wants to Yes… 

Robert:  Give you my card. 

Aleks: I would love to get your card. Yeah. Thank you very much. 

Robert: I’m working at hospital and my email address,

Aleks: Okay, so there is the website. www.iapjapan.org. 

Robert: Yeah. This is the… 

Aleks: Oh, that’s the International Academy of Pathology, Japan. Okay. So definitely we can check this website. 

Robert: This the, next year we host the International Congress.

Aleks: Oh. So that’s a good congratulations in Fukuoka…

Robert: yeah. Fukuoka. That’s the southern part. Big city.

Aleks: Okay. So if anybody wants to visit Japan in Fukuoka, 8th to 13th of October. Thank you so much.

Robert: You’re welcome to come, Yeah.

Aleks: Yes. Now we have to go to Japan.

Robert: Yeah,