Episode 373 - Medical Imaging on Amazon FSxN for NetApp ONTAP

This week on the Tech on Tap podcast, we talk about medical imaging in the cloud with AWS FSXN. Welcome to the Tech on Tap podcast. With Justin Parisi, a love NetApp.

Oh, yeah. Get up! I love this company. Zipok! Zipok! I love NetApp because it's so fun.

Hello and welcome to the second tap podcast. My name is Justin Parisi. I'm here in the basement of my house and with me today, I have a couple of special guests to talk to us all about medical imaging specifically with Amazon FSXN. So to do that today, we have Kim Garriott. So Kim, what do you do and how do I reach you?

I am the general manager for global healthcare and life sciences at MedApp and I have spent over 25 years of my career, helping organizations understand how to have best design clinical workflow and data management strategy around medical imaging and other critical clinical data in our healthcare enterprises.

Awesome. Also with us today, we have Sasha Pagle from AWS. So Sasha, what do you do over there at Amazon and how do we reach you? I work in what's called the Worldwide Specialty Organization, or WSO.

I am directly aligned with our storage services. I focus on healthcare and life sciences, workloads, specifically medical imaging and R&D workloads like genomics,

and really helping customers be able to choose the right collection of storage services along with compute to enable their workloads in the cloud. All right, excellent.

So we're here to talk about medical imaging and we're gonna talk about it in context of Amazon FSXN. So to do that, we wanna first talk about

what medical imaging actually is, what it means, and what sort of things are involved in it. So Kim, give me your overview of medical imaging. Yeah, so medical imaging as we think about it today

is really different than we maybe thought about it 20, 30 years ago. Today when we talk about medical enterprise imaging, it really is looking at all of the different

medical images that are acquired across a healthcare enterprise. Historically, we think of that as being radiology exams, CTs, MRIs, X-rays,

cardiology exams like ECHOs and EKGs. But then there's also a whole world of other imaging when you think about digital photography documentation of post-op wounds or surface anatomy

for lesions on the skin that a dermatologist may track. And now we also see what I like to think of as the final analog imaging workload that we have in healthcare, which is pathology.

So we're now starting to see lots of interest in movement into digitizing the glass pathology slides and now presenting those as digital content for the pathologist to interpret visually

through the computer. And the reason that I've been passionate about medical imaging really for my entire career and why we have a big focus on medical imaging at NetApp

is because of all the data in the world. When you look at the global data sphere, we know that healthcare data comprises approximately 30% of the entire global data sphere.

When we look at healthcare data, we know that 80 to 90% of all healthcare data is comprised of medical imaging. So when we think about the data

that's generated in a hospital, medical imaging is big data. So we're really excited to work to develop solutions that are going to serve healthcare organizations

and customers in the best ways possible in learning how to manage these really large scale data implementations and deployments. It's pretty impressive to see what particular organization

by the year might generate a petabytes worth of digital pathology images. So definitely data's getting bigger. It's getting bigger by different imaging modalities

and there's overall just much more of it. With any sort of large data set, there's always going to be a subset of challenges. So what are the specific challenges that are unique to a medical imaging workload that you might

experience when you're trying to store it either on-prem or in the cloud? Yeah, so number one is really scale. How do you keep up with the capacity to be able to scale quickly enough to be able to have the storage to hold those images or retain those images? And we talked a little bit about digital pathology, and Sasha just made a quick reference

that an organization could easily be generating a petabyte of data a year just from the adoption of a digital pathology workflow. When we look at what we've historically seen with medical imaging, we're more in the terabytes per year generation. And even for smaller organizations, it may be low number of terabytes. But digital pathology in this evolution that we're seeing in the

the field of pathology is going to have a 10 times multiplier on that data. So organizations and CIOs as they're thinking about their long-term IT strategies are really focusing in on how do I most effectively scale to meet the growing demand and the growing capacity of the medical imaging data?

Along with that, the equipment vendors are always improving the technology. If you look back to the late 80s when we first introduced MRI technology, we may have only had a handful of images that makes up that study. Well now you fast forward and we can have thousands of images that make up a single exam

that you may have or MRI that you've had taken as a patient. So the technology is increasing the capacity and the volume of data being generated as well. So these things are very top of mind for organizations as to how they're going to manage the

data. Couple with that a need to really achieve operational efficiencies. We're still recovering from the pandemic in health care as most industries are, but very much so our hospitals and health systems were hit very hard. And we're still operating in negative margins in 60% of the health care organizations across the United States. So we're looking to understand how we can

and manages these ever increasing volumes of data in the most cost-effective ways, and then layer on cybersecurity and the concerns around data protection,

which those threats only become more rapid and increasing every day. So how do we really protect the data in the best ways