Our Chief Product Officer Nathan Buchbinder is the latest guest on Dell Technologies’ Healthcare PowerChat, a podcast series that brings weekly updates from healthcare industry experts on healthcare technologies and transformation.
In part I of his episode, “Accelerating Digital Pathology (#60),” which originally appeared on Dell Technologies’ SoundCloud page, Nathan starts by describing the challenges of today’s outdated practice of pathology – from pathologists’ increasing workload to the slippery slope of subjective diagnosis. Next, Nathan explains how recent advances are enabling pathology to finally overcome the hurdles of going digital. Nathan shares our strategy for helping laboratories adopt digital pathology – from our Concentriq® software platform for image and workflow management to our AI applications and key infrastructure partnerships, including Dell Technologies and NVIDIA. On the regulatory side, Nathan discusses how Proscia’s Concentriq Dx solution recently received CE mark, enabling its use for primary diagnosis in a clinical setting in Europe and other key geographies globally. Listen to the full episode below.
In part II of his episode, “Accelerating Digital Pathology (#61)”, Nathan Buchbinder discusses our AI applications, including DermAI, and how they are helping to improve laboratory efficiency and confidence. Further on, Nathan describes Proscia’s partnerships with Dell and others in the pathology ecosystem. Finally, Nathan shares his thoughts on where the cloud fits in the management of digital pathology images. (Hint: for many customers, it’s the best way to scale to keep up with the amount of images being produced.) Listen to the full episode below.
For the full story, listen to the podcast, or read the transcription below. And don’t hesitate to reach out if you have any questions!
Transcription (Part I)
Intro: Welcome to the Dell Healthcare PowerChat podcast series where you hear from the experts about healthcare technologies. (0:04 – 0:11)
Bruce Hall, Host: Hello, everyone. Bruce Hall here, and welcome to another Dell Healthcare PowerChat. Today we’re going to get into the world of digital pathology with Dell partner, Proscia. Our guest is Mr. Nathan Buchbinder, who’s the Chief Product Officer and Co-Founder of Proscia. How are you doing today, Nathan? (0:17 – 0:31)
Nathan Buchbinder, Chief Product Officer, Proscia: Good. Thanks for having me, Bruce. (0:32 – 0:33)
Bruce: Could we start with a little bit of your background, please? (0:34 – 0:35)
Nathan: Yeah, absolutely. I’m the Chief Product Officer and a Co-Founder here at Proscia. We’re a digital pathology solutions provider – a software company that builds platform and AI technology for the pathology space. We’re looking to change the way that the world practices pathology, one of the core elements of diagnostic medicine. (0:36 – 0:53)
Bruce: Let’s start to talk about digital pathology. Pathology is known as the last -ology to go digital, yet it’s desperately needed to determine potentially very serious diagnoses. Why the delay to go digital with pathology? (0:54 – 1:06)
Nathan: Yeah, it’s a really good question. It truly is the last medical -ology to go digital largely because it’s nearly 100% analog as a practice. For the past 150 years, 100% of cases have been viewed by the pathologist looking at tissue that’s put onto a glass slide under the microscope. It’s a process that’s inefficient. U.S. pathologist supply can’t keep up with the demand of biopsies that need to be read. A pathologist today reads 40% more cases than they did 10 years ago, and pathology is riddled with subjectivity because of its very manual nature. We see pathology as a $30 billion industry – and growing – that can really leverage digital transformation and augmentation to help the pathologist and ultimately potentially impact the diagnosis and treatment for millions of patients every year. (1:07 – 1:57)
Bruce: So what are the hurdles to digitizing the pathology data that you speak of? (1:58 – 2:02)
Nathan: I’d say there are three real hurdles. The first has been on the technology front. The images that are created in pathology when you digitize pictures of tissue biopsies – they can be well over a gigabyte in size. That requires a huge amount of infrastructure and storage that, until recently, hadn’t been available or wasn’t affordable. It requires a lot of compute infrastructure as well when you start running analytics. The second key hurdle has been from a value creation perspective. Digital pathology has only recently become commercially viable and showing positive return on investment for labs that adopt it, and we’re starting to see that with the AI applications that are driving efficiency and improving quality. And the last real hurdle has been in regulatory where we’ve actually seen quite a few of the traditional roadblocks being overcome by many companies – hardware and software companies alike – both in the U.S. and in Europe. (2:03 – 2:57)
Bruce: So as we evolve to overcome these hurdles, how is the digitization actually happening? (2:58 – 3:03)
Nathan: On the front-end, the starting point for any digitized lab is with the whole slide image scanner that takes a glass slide and creates a massive whole slide image. After you’ve created that whole slide image, you need software. So now we have software that takes these digitized images and creates a digital ecosystem – a single workspace for pathologists, researchers, and technicians to actually use those images to move them through a process that’s guided by software. Regulatory agencies have started to take a look at the hardware and software side of this, and we’re starting to see more favorable climates, both from the regulatory perspective with CE mark that allows for primary diagnosis in a clinical setting in Europe and other geographies outside of the U.S., as well as from a value creation perspective once you’ve entered this digital realm. (3:04 – 3:52)
Bruce: So we want to create this digital ecosystem for digital pathology. Let’s talk about Proscia. What is Proscia’s strategy for supporting this digital transformation? (3:53 – 4:02)
Nathan: Proscia is a software solutions provider with both a platform to support the pathologist or researcher after the whole slide image has been created as well as AI applications that layer on top that launch from our platform to help drive efficiency and quality and ultimately augment what the pathologist is capable of. So from a platform perspective, Proscia offers a product called Concentriq. It’s our digital pathology platform to help manage workflows. So once you’ve digitized an image, you have to ask the question of what’s possible with that image. Our platform Concentriq is what enables the image to be a part of a centralized workspace for pathologists – for them to go through case review, for them to share these cases in real time with their colleagues, and get access to expertise from anywhere in the world. Ultimately, that same workflow platform Concentriq serves as a means of implementing new AI tools and applications that can help drive efficiency in a way that you can’t do with physical glass. (4:03 – 5:03)
Bruce: So the platform is Concentriq. Tell us how Concentriq addresses some of the hurdles you mentioned earlier on the journey to digital pathology. (5:04 – 5:10)
Nathan: We can help with managing the images themselves with key infrastructure partnerships and solutions that allow us to help laboratories scale up their digital implementation and help pathologists manage their workloads. With AI applications, you add more value so that easy cases can be pre-sorted and classified prior to a pathologist’s review. All of this takes a fair amount of software. We’ve been lucky enough to partner with Dell Technologies and with NVIDIA to create reference architectures that are scalable and make this something that’s actually not just deployable, but a real application and a real driver of growth in the pathology space. We’re working with Dell Technologies and NVIDIA to push the boundaries of what’s possible with digital pathology today. I should note, by the way, we actually, on the regulatory front, just got CE mark for our platform Concentriq, so it’s just another step forward in the broader digital adoption that we’re able to help labs provide to their pathologists. (5:11 – 6:09)
Bruce: Congratulations on that CE mark, and I know it’s something you were anticipating. And that was just recently issued, is that correct? (6:10 – 6:15)
Nathan: That’s right. That was just in October of 2019, so we’re thrilled to have been able to get that through. What it means to us is that we’re producing not just safe, but high-quality products that really can empower a clinical workflow. We’re excited to go beyond where we are now both in the platform and with our AI technologies as well. (6:16 – 6:37)
Outro: This concludes part I of our podcast. Check back for part II where Nathan discusses the AI components of Concentriq and DermAI™ in particular, how AI actually means ‘augmented intelligence,’ not ‘artificial intelligence,’ Proscia’s partnerships with Dell and others in the digital pathology ecosystem, where cloud fits into the picture, where to find more information, and final thoughts. (6:38 – 6:59)
Transcription (Part II)
Intro: Welcome to the Dell Healthcare PowerChat podcast series where you hear from the experts about healthcare technologies. (0:05 – 0:11)
Bruce Hall, Host: Welcome to part two of our podcast on digital pathology with Dell partner Proscia and our guest Nathan Buchbinder, who’s the Chief Product Officer and Co-Founder of Proscia. In part one, Nathan first discussed the hurdles that have prevented pathology from going digital, how digitization of pathology images actually happens from hardware, software, and regulatory perspectives, Proscia’s Concentriq software platform for managing digital pathology data and workflows, and Proscia’s recent grant of the CE Mark for its Concentriq Dx solution. Let’s pick up the conversation. Nathan, you mentioned the AI component. When I think of looking at one of these slides, it’s this visual chaos of a whole mess of information. Can you describe how your AI component of Concentriq augments your digital image workflow? (0:16 – 1:00)
Nathan Buchbinder, Chief Product Officer, Proscia: You’re 100% right. It really is a big mess of visual chaos on each of these slides. When you think about what a pathologist is doing when they’re looking at these slides, they’re trained to recognize patterns through medical school, through residency, through fellowships, and through years of experience. In recognizing patterns, they’re able to identify what the diagnosis is and what the likely outcome is for the patient and whether the patient is going to respond to certain therapies or not. These are really important, critical questions. We took an approach where we determined if there are any of these patterns that AI can actually recognize, that the AI could detect in a way similar to or potentially unique from how the pathologist is looking at these cases, and then ultimately provide that information to the pathologist in a way that helps them, that allows them to move faster and more confidently through these cases, that makes sure they don’t miss anything, that gives them more information as they’re going through the critical task of looking at a patient’s biopsy and giving a result.
It might be helpful if I gave an example of one of our applications, DermAI™, which is an algorithm for skin pathology that’s able to look at any skin biopsy case and determine what pattern it fits into and what category of case it is. Is it going to be a high impact one, something that looks more like a melanoma, or is this going to be a lower impact one, one that’s a little bit easier for the pathologist to work through. By having that information in advance, we can actually provide it to the pathologist in a way that helps him move through cases faster, look at the cases that matter to them a little bit more and have them focus attention on the cases that ultimately require the most in-depth understanding. So we’re really focused on optimizing that interface between the pathologist and between the laboratory and how it traditionally runs and the digital image, the platform itself through the use of AI as a powerful augmentative application. (1:01 – 2:57)
Bruce: And you said something that we’ve heard before on the Healthcare PowerChat podcast, which is this idea of augmented intelligence as opposed to artificial intelligence. We’re not trying to replace the provider. We’re trying to augment their information. Is that correct? (2:58 – 3:11)
Nathan: That’s 100% right. The role of the pathologist today is not just to serve as a pattern recognition device. The role of the pathologist is to serve as the human being who’s looking at these biopsies and then ultimately translating that into something that’s meaningful — translating that into something that can be related to the oncologist or related to the specialist who sent the biopsy down, who’s working with patients and translating that into answers to the questions of who to treat and how to treat them and if those treatments work. It’s critical that we give these pathologists as much as we possibly can in terms of additional information. So this isn’t about replacement at all. This is about helping pathologists work at their peak, helping them perform as best as they can and focus on the elements of their work that matter the most to patients. (3:12 – 3:57)
Bruce: You mentioned that you have a software solution and that you create an ecosystem and you have key partners. Let’s talk about those partners, and to begin with, can you describe your partnership with Dell? (3:58 – 4:07)
Nathan: It’s been incredible to work with you guys over at Dell. We’ve been able to quickly scale out an infrastructure framework, reference architectures that help us to deploy to some of the largest enterprises in the world. We needed something like this because ultimately what we’re talking about is a major transformative technology that takes a team, that takes efforts from the software side to build a scalable workspace, a scalable ecosystem for the pathologists and the researchers to do their work, but also takes that scalable enterprise-class infrastructure solution that Dell and the partners that Dell has helped us to connect with — groups like NVIDIA, for example — who we work with to create a powerful back-end that supports this drive into the future. (4:08 – 4:53)
Bruce: Inevitably, when we think about digitizing pathology images and slides, you’ve mentioned yourself we could have one image well over one gigabyte in size. We’re talking about a lot of data here. Where does cloud fit into the management of these images? (4:53 – 5:07)
Nathan: It’s interesting. From day one, we’ve always seen cloud, for many customers, as the only way that they can scale up with the amount of data that’s being produced, and through working with organizations like Dell through these kinds of collaborations between the software and the infrastructure itself, we’re able to leverage solutions like Dell’s cloud services, which provide a platform for us to use Amazon Web Services or Azure or other algorithms and launch that in a manner that’s best for the customer. So whether they’re looking for a cloud solution, an on-premises deployment, or something that’s a hybrid, it’s really important that the application itself – the digital pathology platform and the software components of it – really tie in seamlessly to the most efficient and most cost-effective architecture. That’s been one of the beautiful things we’ve been able to put in place with Dell. (5:08 – 6:01)
Bruce: Again, you mentioned an ecosystem. What other partners besides Dell do you work with? (6:02 – 6:05)
Nathan: In addition to working with Dell, we’ve got a hardware-agnostic platform. That means you can use any whole side image scanner and ingest images, so that means Huron Digital Pathology, which produces a great scanner, Leica, Hamamatsu, Epredia, which distributes the 3DHISTECH scanner. There’s an endless list of them. We support all these image formats. On the other side of this, we’re also working with LIS vendors, groups like XIFIN LIS, which provide the information management outside of the images of the slides themselves. And through supporting this broader ecosystem of software, hardware, and image analysis vendors, we’re really centralizing the work of the pathologist. We’re making it a no brainer for these labs to go digital with us. (6:06 – 6:53)
Bruce: Right. Nathan Buchbinder, Chief Product Officer and Co-Founder of Proscia. Where can listeners find more information on Concentriq and the rest of your solutions? (6:54 – 7:00)
Nathan: You can actually learn a lot more from our website. You can find us at Proscia.com, so that’s P-R-O-S-C-I-A.com. We’re always happy to pick up a call from you and learn what you guys are looking for, so I encourage anybody who’s listening to reach out. (7:01 – 7:16)
Bruce: Nathan, thank you so much for all this great information on digital pathology and how far we’ve come in this practice. Any final thoughts? (7:17 – 7:22)
Nathan: No, Bruce, thank you so much. It’s been an absolute pleasure to lay this groundwork with Dell, and we’re excited to see how we continue to scale this out to serve an increasingly large group of labs that are going digital and joining this transformation. So thanks to you and the team so much for working with us on this amazing opportunity. (7:23 – 7:41)
Bruce: Well, that’s our pleasure, too, Nathan, and hopefully we can have you back on a future Dell Healthcare PowerChat. (7:42 – 7:46)
Nathan: Thank you so much, Bruce. (7:47 – 7:49)
*In the U.S., Concentriq Dx is not cleared by the FDA for use in primary diagnosis.
New to digital pathology? Learn how whole slide imaging is changing the industry here.