Red Hat Summit 2024 | Dr. Ellen Grant & Dr. Rudolph Pienaar | Boston Children's Hospital

Clips
News
More from Red Hat Summit 2024

Dr. Ellen Grant

Director, Fetal Neonatal Neuroimaging and Developmental Science at Harvard University

Boston Children's Hospital

Dr. Rudolph Pienaar

Assistant Professor in Radiology

Boston Children's Hospital

Boston Children’s Hospital leverages AI-assisted diagnosis with Red Hat OpenShift AI

AI-assisted diagnosis is transforming healthcare by easing the cognitive load on medical professionals and enhancing decision-making.At Boston Children’s Hospital, the use of Red Hat OpenShift AI in maternal fetal healthcare is providing more accurate diagnostics and streamlined treatments. This innovative technology promises to reduce physician burnout and improve patient outcomes, signifying the impact of AI in the medical field, according to Ellen Grant (pictured, left), director of faculty affairs for the Department of Radiology and director of research, Maternal Fetal Care Center, at Boston Children’s Hospital

play_circle_outline AI's potential in the medical field

play_circle_outline Medical applications of AI in maternal fetal healthcare

play_circle_outline Use of AI to reconstruct fetal brain images and reduce cognitive load for radiologists

play_circle_outline The Revolutionizing Power of AI and Open Source: Implications for Healthcare Delivery and Data Privacy in Medical Imaging

Info

Dr. Ellen Grant & Dr. Rudolph Pienaar | Boston Children's Hospital

Dr. Ellen Grant

Director, Fetal Neonatal Neuroimaging and Developmental Science at Harvard University Boston Children's Hospital

Dr. Rudolph Pienaar

Assistant Professor in Radiology Boston Children's Hospital

search

Rebecca Knight

>> Good afternoon everyone, and welcome back to theCUBE's live coverage of the Red Hat Summit here in beautiful, sunny Denver, Colorado. I'm your host, Rebecca Knight, along with my co-host and analyst, Rob Strechay. When I think about the promise and power of AI, the biggest use case that attracts me and interests me is in the medical field.

>> Yeah. I look at it and go, there is so much that you could do. And what we hear about, and it's traditional AI, it's not the new-fangled GenAI.

Rebecca Knight

>> It's not even the new stuff, exactly.

>> It's what people have been doing with ML for years. And I think that's what excites me is there's a lot of history with this and it's just going faster now.

Rebecca Knight

>> Yeah, there's a lot you can do with the old-fashioned inference stuff. So a perfect segue to our next guest. I'd like to welcome Dr. Ellen Grant and Dr. Rudolf Pienaar. They are both from Boston Children's Hospital and Harvard Medical School. Thank you both so much for coming on the show.

Dr. Ellen Grant

>> Thank you.

Rebecca Knight

>> Returning to the show. I know you're both CUBE alums. So I think a good place to start is to talk about how AI is used in clinical settings because even though there is so much promise and potential, it's actually not used that much. Can you talk a little bit about the lay of the land and where we are now, Dr. Grant?

Dr. Ellen Grant

>> Yeah, sure. Typically in radiology, it's not used that often once the images is at our table. There's some AI baked into image acquisition, so there's deep learning acquisitions. So it's in the sequence acquisition, for example, in MR and for some denoising, but by the time it gets to us do the images, we just read them still. So there's not much more. Other than that, it's all visual perception and interpreting the images, and there's very little AI that happens to the image once it gets to our hands. And the reason for that is it's really hard to then take AI algorithms and put them into our clinical workflow. So there's a lot of people jostling to have access to provide us that analysis before it gets in our hands. But very few things actually have made it.

Rebecca Knight

>> Okay, okay. And so how is Boston Children's Hospital using AI right now to make decisions?

Dr. Ellen Grant

>> Well, right now we're developing tools to be able to improve our ability to do decision making, and we're focusing mostly on the image analysis part in terms of physician-assisted diagnosis. So AI is not making decisions for us. We're trying to get AI to simplify things for us, like decrease the cognitive load was a word I heard this morning, which I really resonate with that concept. And to also give us quantitative data before we were just looking at something. So to give us numbers, because if we could measure it, we can look at how we can improve outcomes or how different treatments have an impact. Whereas qualitative assessments are much less precise.

Rebecca Knight

>> Okay.

>> Yeah, I think what's really interesting is that, and we've talked about this a little bit leading up to this, is that we're at basically an open source conference and Red Hat, the foundation of Red Hat is open source. Really when you look at it, medicine as we were talking about, makes a lot of sense from an open source perspective. How do you both see it from working with Red Hat, open source, medicine, working with other hospitals for that matter. Because again, that becomes where it can be really exciting.

Dr. Ellen Grant

>> Yeah, no, absolutely. Do you want to start with that one?

Rebecca Knight

>> So I would say it's fundamental to how we see things. It's open source, open science. We're not in the business to try and be proprietary about the things we do. That's not what it's about. We're about getting things out there to help the most, the maximum number of people possible. So the link with Red Hat is completely natural because if we develop tech that runs on a platform, I want everyone to use it, without a doubt. But medicine is a strange environment as well from a tech perspective because in some ways it's very advanced, but in some ways from a computing perspective, it's very 1990s. So a lot of stuff doesn't happen, but a lot of stuff is about to happen. And I quite feel quite passionately that if we can get in with an open source approach to distributing this stuff, we can have a new way of doing things that isn't the old way where vendors come in, each one sells their particular product, they don't work together, they don't talk to each other, and we don't advance the field forward. So I think open source really levels that completely.

Dr. Ellen Grant

>> And what I really like about working with Red Hat and OpenShift for example, is we can create our own platform, which is ChRis on OpenShift, and then we can give it to somebody else who's running OpenShift and it runs out of the box. When we are trying to take algorithms from other sites, even if they're containerized, by the time we bring them to our site and try to run them, that's two, three weeks of a highly sophisticated team's work to get it running. So there's so much overhead to take other people's tools. So if we put it in a ChRis environment on OpenShift, it becomes easily shareable to anybody that's running OpenShift and as we were talking about earlier today. And then if we put it in an open cloud like the mass open cloud, then people who don't have compute resources on-prem have access. So that what really excites me is to be able to put more algorithms inside ChRis so everybody can access to it and for everybody it runs in the same manner. So we have our favorite tools, but I'd love other people to put their tools in as well so we can have a community that actually shares those resources in a platform that makes it easy to do that last mile. Something we've been passionate about for years is this ChRis platform. It doesn't end up on my computer or my research office, it ends up in the workstation where I'm reading the cases. And if it doesn't do that, it's not going to make it into clinical practice.

>> And what is ChRis, because you've said it now. I'm always big and I always get it called out for using acronyms, so I'm going to call you guys out for that. What does ChRis stand for?

>> So it used to stand for the Children's Research Integration System. But as time went on, it just became recursive and it became the ChRis Research Integration system because I'm kind of nerdy, so that tracks.

>> It's good.

>> But what it really is, it's something that sits on top of OpenShift and brings a more kind of desktop type experience to running your apps. That's probably the best way to describe it. It makes it very easy for researchers to get their programs on ChRis, and then on ChRis it can go into OpenShift. And they don't have to know anything about the cloud or anything like that. So that's a big win.

Rebecca Knight

>> So both of you were up on the main stage this morning at the Red Hat Summit talking about the strides you've made in maternal fetal healthcare using Red Hat OpenShift for artificial intelligence. Can you tell our viewers a little bit about the progress that's being made?

Dr. Ellen Grant

>> Well, two of the favorite projects we were talking about today that we think are going to have at a big impact. One is when we're imaging fetuses, and we do MRI on fetuses because one of the most accurate diagnostic modalities for fetal health. So once often we look at fetal brains, for example, and it's really hard because the fetuses are moving. We can't tell them to stay still. So they're moving around swimming around in the uterus as we're taking the images. So as you're taking images, what should be just straight images through a head like this, they're all over the place and different angles, and some are blurry because they've moved partly through the acquisition.

Rebecca Knight

>> Those babies.

Dr. Ellen Grant

>> Yeah, I know. If they'd only stay still. So you end up then we keep repeating and repeating and repeating, try to get that perfect stack of images. So at the end, I'm sitting there with maybe a thousand images I have to look at and try to reconstruct that brain in my mind and make sure I see every piece of it. And I'm spending so much time trying to reconstruct that brain. Now what we've done in collaboration with MIT is developed an AI algorithm that works on OpenShift, that takes all those images and reconstructs them and sort of shuffles them into place and gives me one coherent volume. And then I could just look at that coherent volume and quickly look at the brain like I would a brain after you're delivered. We can strap people down and hold them still. So this one gives me a coherent volume that I can just read very nicely so I can put my brain power on the diagnostic component of it instead of the exhausting trying to put this back together in my head instead of having an algorithm do it. So that's one algorithm. And then once we integrate it into the ChRis platform, we can put it right at the bedside. So we've got listeners in the ChRis platform, so it listens for those images to be dumped into our PACS archive or image archive, grabs them, runs the algorithm, puts it back on my desktop within five minutes. So we're just about to deploy that, and I'm really excited to see the impact it has on cognitive load and the accuracy of our diagnoses. I think it's going to be so much easier and less exhausting to use a whole volume.

Rebecca Knight

>> Because the cognitive load is real.

Dr. Ellen Grant

>> Yes, yes.

Rebecca Knight

>> I mean, we know so much, particularly in the aftermath of the pandemic about burnout in the healthcare industry. And so if we can reduce that, it really could change things.

Dr. Ellen Grant

>> And it gets to the point in Geoffrey Hinton's comment about I would be out of a job by now is that we're not anywhere near radiologists being replaced. Where we're trying to get to is that cognitive load. Because we are short staff, we work long days and we barely look up from the computer screen. We really need help in just sorting things, making things easier. It's going to be a long time until the computer and AI algorithm can make a clinical medical decision for us. But there's a lot that can be done to make it easier for us. So we really focus on what we want to do, which is the diagnostic component.

>> And this is also deployed, whereas part of the Massachusetts Open Cloud as well.

Dr. Ellen Grant

>> Yes, it'll work on the open cloud as well too. But what we're working towards is wanting to get that HIPAA compliant. Because right now we can anonymize research data and put it up there temporarily, but if we really want a platform there that anybody could access, we need a HIPAA compliant environment. So that's what we're working on to try to get that set up so we can start to then deploy to local hospitals or even move into LMICs, the low and middle income countries that only have the image acquisition devices, but no compute power whatsoever. So being able to bring that to anybody is something we're really excited about too.

>> Yeah, I've actually heard that from a radiology perspective, especially on the pharmaceutical side where they're taking different pictures from different angles from different machines and how, if they have to move it from one country to another country and how they have to anonymize the data. And is that still, that would seem, I mean again, you're Boston Children's Hospital, so you don't necessarily have that kind of thing, but when you're sharing the data you do. So is that where a lot of this, how do you share the data and is that how everybody's kind of working together to come try to look at formats and things of that nature?

Dr. Ellen Grant

>> Yeah. We're looking at imaging, we have a DICOM format, so there's more of a standardization around images for most medical imaging. And then the other thing about the open cloud and being involved in that whole research environment is the potential to start to get the multi-party compute. So everybody has their own enclave within the cloud that you keep your own data private, but any algorithm can compute on any of the data. So that's hopefully where we'll get to. We still can share with other sites and institutions with special data use agreements, and then we share it in anonymized fashion. I still think we can move forward in that manner. Many projects are going on in that way, but it would be better to go to that next level and have everybody have their own environment so that we don't have to worry about all these data use agreements.

Rebecca Knight

>> How do you foresee this changing the culture of the hospital and in research in general? I mean, something you had said earlier, Rudolph, about how you describe this moment of almost a precipice that a lot has happened but a lot still needs to happen and could happen. You see all this promise and potential. We're not quite there yet, but we're really almost there. How do you foresee this really revolutionizing the way people do their jobs and also the way healthcare is delivered to us patients?

Rebecca Knight

>> Oh, completely. I think my day gets filled with a lot of drudgery. Anyone's day is. And if we can eliminate a lot of that, it begins to make a huge difference. You were just talking about the DICOM anonymization, that's drudgery work. There's nothing sexy, there's nothing fancy about doing that. And people hate to do it because there are no tools that make it easy to do. So part of what we try to do is to make even the drudgery parts really easy because that's how you break down the barriers. In my experience, if someone needs to do anything beyond their normal workflow, they're not going to do it, not easily. It has to be drop dead simple. So to do these kind of things, to have anonymization be in the platform will make hospitals or researchers much more able to be willing to contribute their data, especially under the HIPAA compliances and stuff like this. So I think from our perspective in medical imaging and data, we're very concerned about data provenance, data privacy. So I want to say you can trust that we would take care of the data. We're not going to use stuff in models that we don't know where it comes from, that you don't know where it comes from. That's what we bake into our systems as well. And these kind of ideas, I think really are now coming together and are going to just open the next level of floodgates, where we can innovate at scales that we don't do today. And that's going to be lead to a lot of exciting innovation.

Dr. Ellen Grant

>> And that's another thing about say the Mass Open Cloud that we really like is the data governance structure. It's not in the hands of some commercial entity, it's in the hands of the academic institutions. And that makes me feel much more comfortable, and we think that's the way to go in terms of medical imaging that can actually support hospitals.

Rebecca Knight

>> How much are you in touch with other people like you who are both practitioners, but also technologists too? I mean, I know we're here at the Red Hat Summit and there's not a lot of people who are in the healthcare field, but how much are you sort of bringing this open source philosophy to other people and talking to colleagues at other institutions?

Dr. Ellen Grant

>> That's a great question. There are very few of us, unfortunately, but I think there are people all over the place. And I think there's a recognition coming that the importance of technology in medicine, especially computing technology, is really where it's at. A lot of medicine is still thinking in terms of pharmacology. That's 20th century. We're in the 21st century, computing is what's going to break the new frontiers. And that idea is beginning to come through. Forums like this, like theCUBE, getting the word out. This all contributes to getting that information out there.

Rebecca Knight

>> Excellent. Well, a terrific note to end on both. Thank you both so much. Dr. Ellen Grant and Rudolf Pienaar. Thank you so much.

Dr. Ellen Grant

>> Great, thank you.

Rebecca Knight

>> I'm Rebecca Knight for Rob Strechay. We are done for the day, but come back tomorrow for more of theCUBE's wall-to-wall coverage of the Red Hat Summit. You are watching theCUBE, the leader in enterprise tech news.

Dr. Ellen Grant & Dr. Rudolph Pienaar | Boston Children's Hospital

search

Rebecca Knight

>> Good afternoon everyone, and welcome back to theCUBE's live coverage of the Red Hat Summit here in beautiful, sunny Denver, Colorado. I'm your host, Rebecca Knight, along with my co-host and analyst, Rob Strechay. When I think about the promise and power of AI, the biggest use case that attracts me and interests me is in the medical field.

>> Yeah. I look at it and go, there is so much that you could do. And what we hear about, and it's traditional AI, it's not the new-fangled GenAI.

Rebecca Knight

>> It's not even the new stuff, exactly.

>> It's what people have been doing with ML for years. And I think that's what excites me is there's a lot of history with this and it's just going faster now.

Rebecca Knight

>> Yeah, there's a lot you can do with the old-fashioned inference stuff. So a perfect segue to our next guest. I'd like to welcome Dr. Ellen Grant and Dr. Rudolf Pienaar. They are both from Boston Children's Hospital and Harvard Medical School. Thank you both so much for coming on the show.

Dr. Ellen Grant

>> Thank you.

Rebecca Knight

>> Returning to the show. I know you're both CUBE alums. So I think a good place to start is to talk about how AI is used in clinical settings because even though there is so much promise and potential, it's actually not used that much. Can you talk a little bit about the lay of the land and where we are now, Dr. Grant?

Dr. Ellen Grant

>> Yeah, sure. Typically in radiology, it's not used that often once the images is at our table. There's some AI baked into image acquisition, so there's deep learning acquisitions. So it's in the sequence acquisition, for example, in MR and for some denoising, but by the time it gets to us do the images, we just read them still. So there's not much more. Other than that, it's all visual perception and interpreting the images, and there's very little AI that happens to the image once it gets to our hands. And the reason for that is it's really hard to then take AI algorithms and put them into our clinical workflow. So there's a lot of people jostling to have access to provide us that analysis before it gets in our hands. But very few things actually have made it.

Rebecca Knight

>> Okay, okay. And so how is Boston Children's Hospital using AI right now to make decisions?

Dr. Ellen Grant

>> Well, right now we're developing tools to be able to improve our ability to do decision making, and we're focusing mostly on the image analysis part in terms of physician-assisted diagnosis. So AI is not making decisions for us. We're trying to get AI to simplify things for us, like decrease the cognitive load was a word I heard this morning, which I really resonate with that concept. And to also give us quantitative data before we were just looking at something. So to give us numbers, because if we could measure it, we can look at how we can improve outcomes or how different treatments have an impact. Whereas qualitative assessments are much less precise.

Rebecca Knight

>> Okay.

>> Yeah, I think what's really interesting is that, and we've talked about this a little bit leading up to this, is that we're at basically an open source conference and Red Hat, the foundation of Red Hat is open source. Really when you look at it, medicine as we were talking about, makes a lot of sense from an open source perspective. How do you both see it from working with Red Hat, open source, medicine, working with other hospitals for that matter. Because again, that becomes where it can be really exciting.

Dr. Ellen Grant

>> Yeah, no, absolutely. Do you want to start with that one?

Rebecca Knight

>> So I would say it's fundamental to how we see things. It's open source, open science. We're not in the business to try and be proprietary about the things we do. That's not what it's about. We're about getting things out there to help the most, the maximum number of people possible. So the link with Red Hat is completely natural because if we develop tech that runs on a platform, I want everyone to use it, without a doubt. But medicine is a strange environment as well from a tech perspective because in some ways it's very advanced, but in some ways from a computing perspective, it's very 1990s. So a lot of stuff doesn't happen, but a lot of stuff is about to happen. And I quite feel quite passionately that if we can get in with an open source approach to distributing this stuff, we can have a new way of doing things that isn't the old way where vendors come in, each one sells their particular product, they don't work together, they don't talk to each other, and we don't advance the field forward. So I think open source really levels that completely.

Dr. Ellen Grant

>> And what I really like about working with Red Hat and OpenShift for example, is we can create our own platform, which is ChRis on OpenShift, and then we can give it to somebody else who's running OpenShift and it runs out of the box. When we are trying to take algorithms from other sites, even if they're containerized, by the time we bring them to our site and try to run them, that's two, three weeks of a highly sophisticated team's work to get it running. So there's so much overhead to take other people's tools. So if we put it in a ChRis environment on OpenShift, it becomes easily shareable to anybody that's running OpenShift and as we were talking about earlier today. And then if we put it in an open cloud like the mass open cloud, then people who don't have compute resources on-prem have access. So that what really excites me is to be able to put more algorithms inside ChRis so everybody can access to it and for everybody it runs in the same manner. So we have our favorite tools, but I'd love other people to put their tools in as well so we can have a community that actually shares those resources in a platform that makes it easy to do that last mile. Something we've been passionate about for years is this ChRis platform. It doesn't end up on my computer or my research office, it ends up in the workstation where I'm reading the cases. And if it doesn't do that, it's not going to make it into clinical practice.

>> And what is ChRis, because you've said it now. I'm always big and I always get it called out for using acronyms, so I'm going to call you guys out for that. What does ChRis stand for?

>> So it used to stand for the Children's Research Integration System. But as time went on, it just became recursive and it became the ChRis Research Integration system because I'm kind of nerdy, so that tracks.

>> It's good.

>> But what it really is, it's something that sits on top of OpenShift and brings a more kind of desktop type experience to running your apps. That's probably the best way to describe it. It makes it very easy for researchers to get their programs on ChRis, and then on ChRis it can go into OpenShift. And they don't have to know anything about the cloud or anything like that. So that's a big win.

Rebecca Knight

>> So both of you were up on the main stage this morning at the Red Hat Summit talking about the strides you've made in maternal fetal healthcare using Red Hat OpenShift for artificial intelligence. Can you tell our viewers a little bit about the progress that's being made?

Dr. Ellen Grant

>> Well, two of the favorite projects we were talking about today that we think are going to have at a big impact. One is when we're imaging fetuses, and we do MRI on fetuses because one of the most accurate diagnostic modalities for fetal health. So once often we look at fetal brains, for example, and it's really hard because the fetuses are moving. We can't tell them to stay still. So they're moving around swimming around in the uterus as we're taking the images. So as you're taking images, what should be just straight images through a head like this, they're all over the place and different angles, and some are blurry because they've moved partly through the acquisition.

Rebecca Knight

>> Those babies.

Dr. Ellen Grant

>> Yeah, I know. If they'd only stay still. So you end up then we keep repeating and repeating and repeating, try to get that perfect stack of images. So at the end, I'm sitting there with maybe a thousand images I have to look at and try to reconstruct that brain in my mind and make sure I see every piece of it. And I'm spending so much time trying to reconstruct that brain. Now what we've done in collaboration with MIT is developed an AI algorithm that works on OpenShift, that takes all those images and reconstructs them and sort of shuffles them into place and gives me one coherent volume. And then I could just look at that coherent volume and quickly look at the brain like I would a brain after you're delivered. We can strap people down and hold them still. So this one gives me a coherent volume that I can just read very nicely so I can put my brain power on the diagnostic component of it instead of the exhausting trying to put this back together in my head instead of having an algorithm do it. So that's one algorithm. And then once we integrate it into the ChRis platform, we can put it right at the bedside. So we've got listeners in the ChRis platform, so it listens for those images to be dumped into our PACS archive or image archive, grabs them, runs the algorithm, puts it back on my desktop within five minutes. So we're just about to deploy that, and I'm really excited to see the impact it has on cognitive load and the accuracy of our diagnoses. I think it's going to be so much easier and less exhausting to use a whole volume.

Rebecca Knight

>> Because the cognitive load is real.

Dr. Ellen Grant

>> Yes, yes.

Rebecca Knight

>> I mean, we know so much, particularly in the aftermath of the pandemic about burnout in the healthcare industry. And so if we can reduce that, it really could change things.

Dr. Ellen Grant

>> And it gets to the point in Geoffrey Hinton's comment about I would be out of a job by now is that we're not anywhere near radiologists being replaced. Where we're trying to get to is that cognitive load. Because we are short staff, we work long days and we barely look up from the computer screen. We really need help in just sorting things, making things easier. It's going to be a long time until the computer and AI algorithm can make a clinical medical decision for us. But there's a lot that can be done to make it easier for us. So we really focus on what we want to do, which is the diagnostic component.

>> And this is also deployed, whereas part of the Massachusetts Open Cloud as well.

Dr. Ellen Grant

>> Yes, it'll work on the open cloud as well too. But what we're working towards is wanting to get that HIPAA compliant. Because right now we can anonymize research data and put it up there temporarily, but if we really want a platform there that anybody could access, we need a HIPAA compliant environment. So that's what we're working on to try to get that set up so we can start to then deploy to local hospitals or even move into LMICs, the low and middle income countries that only have the image acquisition devices, but no compute power whatsoever. So being able to bring that to anybody is something we're really excited about too.

>> Yeah, I've actually heard that from a radiology perspective, especially on the pharmaceutical side where they're taking different pictures from different angles from different machines and how, if they have to move it from one country to another country and how they have to anonymize the data. And is that still, that would seem, I mean again, you're Boston Children's Hospital, so you don't necessarily have that kind of thing, but when you're sharing the data you do. So is that where a lot of this, how do you share the data and is that how everybody's kind of working together to come try to look at formats and things of that nature?

Dr. Ellen Grant

>> Yeah. We're looking at imaging, we have a DICOM format, so there's more of a standardization around images for most medical imaging. And then the other thing about the open cloud and being involved in that whole research environment is the potential to start to get the multi-party compute. So everybody has their own enclave within the cloud that you keep your own data private, but any algorithm can compute on any of the data. So that's hopefully where we'll get to. We still can share with other sites and institutions with special data use agreements, and then we share it in anonymized fashion. I still think we can move forward in that manner. Many projects are going on in that way, but it would be better to go to that next level and have everybody have their own environment so that we don't have to worry about all these data use agreements.

Rebecca Knight

>> How do you foresee this changing the culture of the hospital and in research in general? I mean, something you had said earlier, Rudolph, about how you describe this moment of almost a precipice that a lot has happened but a lot still needs to happen and could happen. You see all this promise and potential. We're not quite there yet, but we're really almost there. How do you foresee this really revolutionizing the way people do their jobs and also the way healthcare is delivered to us patients?

Rebecca Knight

>> Oh, completely. I think my day gets filled with a lot of drudgery. Anyone's day is. And if we can eliminate a lot of that, it begins to make a huge difference. You were just talking about the DICOM anonymization, that's drudgery work. There's nothing sexy, there's nothing fancy about doing that. And people hate to do it because there are no tools that make it easy to do. So part of what we try to do is to make even the drudgery parts really easy because that's how you break down the barriers. In my experience, if someone needs to do anything beyond their normal workflow, they're not going to do it, not easily. It has to be drop dead simple. So to do these kind of things, to have anonymization be in the platform will make hospitals or researchers much more able to be willing to contribute their data, especially under the HIPAA compliances and stuff like this. So I think from our perspective in medical imaging and data, we're very concerned about data provenance, data privacy. So I want to say you can trust that we would take care of the data. We're not going to use stuff in models that we don't know where it comes from, that you don't know where it comes from. That's what we bake into our systems as well. And these kind of ideas, I think really are now coming together and are going to just open the next level of floodgates, where we can innovate at scales that we don't do today. And that's going to be lead to a lot of exciting innovation.

Dr. Ellen Grant

>> And that's another thing about say the Mass Open Cloud that we really like is the data governance structure. It's not in the hands of some commercial entity, it's in the hands of the academic institutions. And that makes me feel much more comfortable, and we think that's the way to go in terms of medical imaging that can actually support hospitals.

Rebecca Knight

>> How much are you in touch with other people like you who are both practitioners, but also technologists too? I mean, I know we're here at the Red Hat Summit and there's not a lot of people who are in the healthcare field, but how much are you sort of bringing this open source philosophy to other people and talking to colleagues at other institutions?

Dr. Ellen Grant

>> That's a great question. There are very few of us, unfortunately, but I think there are people all over the place. And I think there's a recognition coming that the importance of technology in medicine, especially computing technology, is really where it's at. A lot of medicine is still thinking in terms of pharmacology. That's 20th century. We're in the 21st century, computing is what's going to break the new frontiers. And that idea is beginning to come through. Forums like this, like theCUBE, getting the word out. This all contributes to getting that information out there.

Rebecca Knight

>> Excellent. Well, a terrific note to end on both. Thank you both so much. Dr. Ellen Grant and Rudolf Pienaar. Thank you so much.

Dr. Ellen Grant

>> Great, thank you.

Rebecca Knight

>> I'm Rebecca Knight for Rob Strechay. We are done for the day, but come back tomorrow for more of theCUBE's wall-to-wall coverage of the Red Hat Summit. You are watching theCUBE, the leader in enterprise tech news.