Ad 00:40
The Medical Alley Podcast is brought to you by MentorMate. MentorMate empowers healthcare clients to deliver on their mission and transform the human experience through technology. For over 20 years, clients have trusted MentorMate to guide their vision, design innovative products, and build secure solutions while understanding the specific nuances of their industry. MentorMate's global team in the US, Eastern Europe and Latin America helps clients in all sectors of healthcare transform their organizations, from Fortune 500 pharmaceutical companies and commercial payers, to hospital systems, medical device manufacturers and beyond. Learn more at mentormate.com/healthcare.
Frank Jaskulke 01:31
Good morning, good afternoon and good evening to everyone out there in Medical Alley. This is Frank Jaskulke, your host of the Medical Alley Podcast. And today, we're going to continue our conversation that we've had a couple of times this year on the topic of AI in healthcare. Today, I'm joined by Sean Cassidy, who's the CEO and co-founder of Lucem Health. Sean, I really appreciate you joining us today.
Sean Cassidy 01:55
Well, thanks for hosting me, Frank. It's great to be here with you.
Frank Jaskulke 01:59
You know, the place I'd like to start and this, it's a loaded question, I know the answer is probably both yes and no, but is AI overhyped in healthcare?
Sean Cassidy 02:09
Yes and no. Sorry, I couldn't resist.
Frank Jaskulke 02:14
It's fair.
Sean Cassidy 02:16
Yeah. So let's start with the no, which is the easy part. It is absolutely the case that these tools have tremendous promise and have the potential to deliver incredible value. I don't like to use, we don't like to use the word transformational. I think that can potentially be too strong. We are early, but certainly the trajectory that we're on, and what we're seeing is immensely promising. And we are getting a very, very good response from healthcare provider organizations in particular and others to that promise and potential.
Frank Jaskulke 02:57
Indeed.
Sean Cassidy 02:58
And what about the what about the yes side? Well, the yes side of it is that a lot of people think that the robots are upon us, and they're going to transform healthcare tomorrow, right? And everything's going to change overnight, and ChatGPT is going to be in the exam room with doctors and so on. Look, we've all been in healthcare a long time. We know how long it takes to drive change, particularly on the technology side. We're a little bit of ways away from, you know, achieving all of that promise and potential that I mentioned.
Frank Jaskulke 03:33
Yeah. Are there places where you might say, hey, it's had some impact already?
Sean Cassidy 03:40
Yeah, I mean, AI has been used in healthcare broadly for some years now. It's been used on the administrative side of healthcare. It's been used, and it's being used quite pervasively in revenue cycle management. It's being used for scheduling, it's being used to support a lot of the so called back office functions that are vital to ensure that we have a healthy healthcare delivery system, you know, at the level of care workflow and in the exam room. And yeah, and increasingly, we're seeing, you know, more and more tools that are available, or sorry, more and more tools supporting AI, population health and value based care tools. So these are tools that are not necessarily used by clinicians directly, but are used by folks who are tasked with, you know, helping patients get to the right settings of care, remind them to do the things that they're supposed to do, taking medication, show up for visits, all of those kinds of things. And those things have been around for a while. The other thing we should say, Frank, is, you know, this term AI is a pretty broad term, right? And so what is — when do we cross the threshold from a very strong predictive algorithm that's using, you know, say pre-existing statistical techniques, when does that thing become, quote unquote, magical and become machine learning or, you know, or something, you know, something that's sort of more modern and more sophisticated in its orientation.
Frank Jaskulke 05:14
Yeah, that's a really important distinction. I mean, it still boggles my mind that when, when we talk more about the popular culture sentiment of AI, kind of, exemplified by the the large language models, ChatGPT, that only entered the broader public consciousness, you know, November of last year, November 2022. But as you said, these tools in different forms have been used for quite some time and have made an impact. Maybe I'd asked you to back up for a moment because you mentioned that transition, which is both a technology challenge and a people challenge. Maybe introduce Lucem Health for the audience. What does it you guys are up to? What is the company doing?
Sean Cassidy 05:57
So we were started back to the idea for the company originated within Mayo Clinic, say three and a half years ago. And the challenge we were intended to confront was the difficulty that AI researchers and AI developers were having trying to get what were really interesting and innovative tools from the bench to the bedside, right, from research mode into care delivery. And so, you know, we were set up to deliver a general purpose platform that could deploy virtually any kind of clinically oriented AI. I want to underscore that we're focused in the clinical realm, not in the administrative side of healthcare that I mentioned before. What we've learned, as new companies do, is that that's a very, very powerful, you know, set of underlying capabilities that need to exist. I'm talking about how do you connect with the data that the AI needs, right, the fuel in effect that's required to power it? How do you orchestrate the execution of a variety of different kinds of models? How do you interpret the output from those models in a way that's human understandable and human relatable? And then how do you surface those insights into the workflow in ways that stakeholders can really, you know, accept, right, and embrace. So that's the stuff, the general capabilities that we've built. But to cut a long story short, Frank, our focus now is on applying that underlying technology to a very specific area, which is accelerating the detection of disease using these powerful novel new tools, AI machine learning, to look at data in ways that can previous, you know, to examine data in ways that were previously unavailable, and find patients who may be at higher risk for certain diseases, and then bring them gently into the care delivery system in ways that allow doctors to practice medicine the way they've been practicing. And they don't have to change their workflows, right? They're just seeing a different cohort of patients, patients who may have certain risk, risk of diabetes, risk of colorectal cancer when they're overdue for screening colonoscopies, risks of heart disease and other cancers and so forth. That's where we are focused now.
Frank Jaskulke 08:29
Oh, that's, okay, I want to break that up into the the two pieces. The platform itself, when I heard you describe that what that made me think was almost like, if I were developing a medical product, but hospitals or clinics didn't exist, where would I go and deploy that product to actually be able to use it and to benefit patients? We figured out how to do that if it's a device or a drug. But if I were developing an AI algorithm, I've got to get the data from the health system, process it, make it presentable, make it usable, like a bunch of different things that probably needs to be replicated over and over again, that I hear you write that the the underlying capability is to, I guess, structure that in a way that it becomes usable.
Sean Cassidy 09:20
That's right, Frank, your hired. So the you know, maybe an analogy that might be helpful to listeners is this, right? Think of an AI algorithm as an engine. It's a very powerful thing. It's got horsepower. It's got torque, right? It's got six cylinders, eight cylinders. It could be electric, it comes in many forms, right? But that engine is no good to anybody unless it's dropped into a car. And so that's what we do. We're car makers. We work with AI innovators, we work with engine makers, and we've constructed our technology infrastructure in such a way that we can build and deploy a lot of different kinds of cars. I have to be careful not to extend too far because it ultimately breaks down. But but it may be helpful to think about the difference between an engine and the vehicle that surrounds the engine. An algorithm does not deliver better care on its own.
Frank Jaskulke 10:18
So the platform part makes sense. Talk to us a bit about the preventative part, this area of clinical care that you're focusing on. What value or what impact does that bring to healthcare?
Sean Cassidy 10:29
So we are very aware of how much providers are struggling these days. They have resource scarcity issues, their margins are highly compressed, if not negative, it feels like every day, right, we get a new story of the challenges associated with clinical staff shortages, or we get a new report of losses faced by large and small provider organizations. And so we are — our mission is to obviously help deliver better health care, not just in the US, but around the globe. And our partnership with Mayo Clinic, obviously, it's not surprising that you're hearing that, Frank. But being aware of these critical issues that face healthcare providers. Today, we're trying to deliver solutions that increase the clinical and financial yield and productivity from existing resources. That translates into ultimately the ability to deliver more and better care. And it also, frankly, does allow for increased revenue opportunities for providers while delivering that care. So we're very, very careful when we think about what solutions we bring to market as to whether it can sort of confront those two issues head on. That's what's motivating us right now.
Frank Jaskulke 11:54
Right on. And I gotta say, kudos on that. Productivity in health care, I think, is such an important part of driving better costs, better quality and better success as businesses in health care. And so if you guys can help that. I'm on board. You've mentioned a couple of times the partnership with Mayo Clinic. And as I understand it, you've been working in particular with Mayo Clinic Platform. Could you maybe tell our audience a little bit about what the partnership involves, and what kind of ongoing work you're doing?
Sean Cassidy 12:32
Yeah, so we have a multifaceted relationship with Mayo Clinic and with Mayo Clinic Platform. Mayo is an investor in the company. They are a customer of ours. Mayo Clinic Platform really was the originator, as I mentioned before, of the idea for the business in the first place. They are our partner in several ways. So they're our partner on the go to market side. They help us with access to their provider network, the Mayo Clinic Care Network, as well as to other relationships that they are developing. They are also our partner on the technology side, and we work with them to think about how technologies and capabilities that support the development, validation and deployment and effective use of AI can be delivered. So there's quite a lot going on there.
Frank Jaskulke 13:31
Yeah, and I love to hear that, the the idea of health systems working to bring new and innovative technologies out but also to get them more broadly adopted so that it benefits other health systems and patients that might be beyond the catchment area of one system. It's just great to hear that kind of work is being done. And as I understand it, Lucem Health also kind of thinks that way, too. And you have this broader, I think it was called the innovation collaborative, where you're working with other companies. What is the innovation collaborative? What does that involve?
Sean Cassidy 14:07
It's a fancy name for our partner program. You know, we like the term because it does, you know, both of those words are meaningful, right? We are trying to find and promote innovation in AI and ML wherever it lives. And it is inherently the idea of taking really clever data science and turning it into a useful solution. It inherently involves a collaboration. But it's our partner program. We're very proud of the folks that are in the program are working now with 35 plus partners. And just, you know, we've only been around for a couple of years, and at various stages of development. But it's been really, really fun to see, you know, not only the development of our own technology and our own team, but also the development and strengthening of our relationships with our partners.
Frank Jaskulke 15:03
Yeah. And maybe to that end of partners, one area in particular, I think you know, we work with a lot of medical device companies, big community here in Minnesota. And many of them have been looking at digital technologies, looking at AI, integrating it into their products. Are you guys working with medical technology companies? And if so, how do you collaborate with that kind of company?
Sean Cassidy 15:28
Well, a couple of ways. So think about a medical device company, of course, it depends on the device, but that device exists for some reason, right? It exists for a patient with a certain condition, or it's a telemetry device that's trying to collect information to determine whether a patient has a condition and so on, and so on. There are many examples, of course, right? But as you might imagine, Frank, medical device manufacturers, and as an aside, by the way, pharmaceutical companies are very interested in solutions that can help find patients with diseases for which they have life changing stuff, right? Molecules, or technology, hardware, you know, whatever. And so you're right, that these companies are increasingly embedding AI within their devices, or in the supporting infrastructure that collects data from those devices. But they're also interested in how AI can actually help them get more embedded in the care delivery process and again, find patients, which is very much what we're about with accelerating disease detection.
Frank Jaskulke 16:43
Oh, that's fascinating. I mean, the idea that we might simplify and improve the processes of, you know, finding people who have a condition and getting them to the right care more rapidly, I think it's what we've we've all been doing this work for. And it's fascinating to see, you know, AI as a technology enabling that to happen at greater scale with greater impact. But also to hear that there are companies working on the infrastructure, the capabilities to translate the the ideas or the hype into reality that makes a difference. So maybe the last thing I'll ask is just, you guys have done a lot in a short period of time already. What's next?
Sean Cassidy 17:29
Well, we, you know, we're going to continue to launch with our partners more and more of these early disease detection solutions. So that is motivating us. And the other thing is that, and this is through our relationship with Mayo Clinic Platform, there are other providers out there, kind of like Mayo, Mayo, of course, is an N of one. There's really nobody else like Mayo in the world. But there are other providers out there who are trying to do AI related innovation. And they're struggling with the same challenge of getting from the bench to the bedside. So how can we help them, you know, take really promising research, promising early development and transform it into something that can deliver value within their own organizations, and then potentially have applicability outside of the four walls of the organization. So there's kind of the retail side, if you like, you know, bringing solutions to market that we want to, you know, hopefully are appealing to every provider regardless of shape and size. And there's a sort of special part of the business where we're working with provider innovators, who really want to get more leverage and mileage out of the research work that they're doing in AI and ML.
Frank Jaskulke 18:47
That's fascinating. Well, okay, one last question, the very simple one, for organizations out there that want to learn more about what you're up to, where should we direct them?
Sean Cassidy 18:57
Well, we'd love to have you come find us on lucemhealth.com. We have published a fair number of white papers on AI in healthcare, and hopefully the practical and pragmatic use of AI. If you're looking to learn more just about AI generally, you know, not a commercial, come to the site and and check us out. And if you are an AI developer and AI innovator, and you're struggling a little bit with, you know, with how to get your really interesting data science into the clinical workflow at scale, we've got a partner program for you. And we talk to everybody. So come find us, and we'd love to chat.
Frank Jaskulke 19:42
Right on. And folks, we'll make sure that's in the show notes so you can track that down more easily. And Sean, thank you. That was fantastic discussion. I really appreciate you sharing a bit of what Lucem Health is up to.
Sean Cassidy 19:54
Well, it's a pleasure, Frank. Thanks so much.
Frank Jaskulke 19:56
And folks, that's been another episode of the Medical Alley Podcast. If you're not already a subscriber, make sure you get over to medicalalleypodcast.org. Or you can find us on Apple, Spotify, and now on our YouTube channel. And hey do me a favor. Would you share this episode with just one other person? If everyone listening did that we'd help spread this story and so many other important stories coming out of the Medical Alley community further. I'd really appreciate it. Until next time, have a great day.