Episode 26: The Alliance for Artificial Intelligence in Healthcare with Key Opinion Leader Elaine Hamm

You're listening to Bio from the Bayou, featuring stories and industry insights from experts in the bustling biotech scene of New Orleans.

The entire Gulf Coast region is buzzing with excitement, and we're here to bring you frontline access to this vibrant ecosystem direct from NOLA, the city that provides a little lagniappe with everything we do.

Where people come for the science but stay for the food, festivals, and resilient culture.

I'm James Zanewicz, the Chief Business Officer for the Tulane School of Medicine.

And I'm Dr. Elaine Hamm, the Executive in Residence for the Tulane School of Medicine.

And this is Bio from the Bayou. Welcome back to Bio from the Bayou. Today's guest is one of our frequent hosts, Dr. Elaine Hamm. She is, of course, the executive in residence at Tulane University School of Medicine, but more importantly for our topic today, she is also the executive director of the Alliance for Artificial Intelligence in Healthcare, Or the AAIH. The AAIH is an international advocacy organization dedicated to the responsible adoption and application of artificial intelligence, or AI, in healthcare. It represents companies and other organizations that are applying AI, machine learning, and technology along a spectrum of uses that enable drug discovery, make it easier for patients to interact with their healthcare providers, Or spot trends in patient data to create digital twins that enable more personalized and effective medicine. That was a mouthful of info. And I want to take a break. That was a lot. I know. And that didn't—

The podcast is over now. Like, you've already done all of my job, so perfect.

And I didn't even cover everything. I know there's more. So I want to welcome you back to the show, Elaine.

Thank you. I'm happy to be here. And I get to be on the other side of it, so that's exciting.

Well, since I get to ask you the questions, can you give us a little more insight into the AAIH?

Yeah. So when you talk about the healthcare industry, it's actually a really broad group, as you can imagine, because it's everything that touches you, the patient, right? I mean, so that can be, like you mentioned, everything from drug development to hospitals, even down to the smartwatches you wear that tracks your biodata. And what our industry needs, because we are so diverse, is having all of us come together to solve related problems, things that we all struggle with. And the AIH plays a really critical role in bringing together diverse stakeholders from the healthcare industry, as well as technology companies that are making the technologies for the healthcare industry, and regulatory bodies that make sure that we are utilizing AI in a very responsible way. And so, yeah, that's the role of the AIH. AIH is helping to bring, you know, really diverse groups together so that we can all solve problems together.

Why do you feel the organization is becoming more and more important as AI becomes more and more prevalent? I mean, we hear it talked about in the public all the time, but AI is way more than ChatGPT. Why is this organization so necessary at this moment in time?

Yeah, I mean, so I will add that the healthcare industry has really been using AI for quite some time. We didn't necessarily call it AI per se, but you know it's still as technology is growing, it's going to really revolutionize how we actually deliver healthcare to patients. And because of that, there is this huge need to have a coordinated effort to to help set up guidelines, to set standards. And as technologies advance, their applications are just going to become more and more prevalent. And so, even though that this growth can really bring some incredible opportunities, it also has significant challenges too. And so we need to be able to make sure that we're doing this in a effort. But again, like I said, the healthcare industry has been utilizing AI for many years. And because we are this already pretty heavy regulated industry, I think that a lot of other industries that are new to AI can learn quite a bit from us as they try to navigate their own AI applications in their industry.

I think that's interesting because as we've heard talked about in the public lexicon, one of the original AIs was really spellcheck in Word and things like that. You know, they're are very simple that we've been very used to. No one realized that was AI. And that's been in healthcare.

Exactly.

And as I mentioned in the podcast intro, much like there were broad applications of things like spellcheck or things like that, the field of healthcare is very broad. So how can the implementation of AI benefit healthcare, biotech, and medical fields? And can you give a few examples as you talk about this of ways it's already used in the healthcare space?

I mean, you've kind of touched on a couple of them that I think are really— one, been around for quite some time and really have an impact on patients, whether they realize it or not. Wherever AI is the most helpful is where there is large data sets. And the healthcare industry does large data sets really, really well. And so that means a lot of different things too, whether it's genomic data for predicting drug development targets or clinical trial data and things like maybe patients may be more familiar with, like clinical notes that your doctor is taking or even imaging. I mean, those are all things that have benefited from the use of AI. So AI, take medical imaging, for example. you train these algorithms on medical imaging interpretation, and it allows for the doctors to really help enhance the accuracy of diagnosis. And it has these predictive analytics to help identify patients that are at risk for specific diseases. And so that can really help the clinician make better decisions more quickly so that the patient can get really quality healthcare and spend more time with their doctor. And the doctor, you know, isn't spending time having to go through tons of imaging data. As we think about maybe something the patient is less familiar with, as we think about drug development, Insilico Medicine, for example, one of our members of the AAIH, they've really achieved this incredible milestone recently where they brought the first drug discovered and designed by generative AI into phase 2 clinical trials with patients. So what they did first was go through— they had a hypothesis that the drug's target discovery involved training this particular engine on clinical data sets related Hmm. Related to a specific indication of tissue fibrosis. And that led to the identification of a brand new novel target as it related to fibrosis as a disease. And then they could go through with generative AI to help discover where they could target that specific target more specifically. So they were able to generate molecules that met all of these really great specific conditions that made it a great drug. And they found, you know, a handful of molecules that could improve fibrosis and also which is really important, demonstrate really good safety. So this ability to predict this is gonna be a good target, this is gonna be a good drug, and we also know it's gonna be safe is incredible. So they went from start to finish into a clinical trial under 30 months, which is insane. So just so like as a frame of reference, I've been working on a hearing loss drug for longer than my husband and I have been married and we are not in clinical trials. So just the thought that AI could take something that is so long and so complicated and so hard And bring it down to months is really, really exciting.

Well, and you're talking about data. And I know when we look at data, there's been a lot of discussion on how diversity of data is important and how historically in clinical trials in the past, it was very homogeneous as far as patient population. So what do you see as the importance of data diversity and access as it relates to creation of AI-enabled future?

Yeah, preventing bias is incredibly important as it relates to AI. And we're, as an industry, really well aware of how we have built datasets in the past that included this inherent bias. I mean, you know, you mentioned clinical trial data. I mean, a lot of it was on, you know, white guys, and that doesn't necessarily tell you whether a drug is going to work in an African American woman. And so the more diverse dataset we have, the better we can predict whether a drug is going to be safe and useful across an entire population, not just a specific one. Even at a research level, I mean, we were doing experiments just on male mice and that the sex differences we are learning more and more is such a huge component of whether a drug is gonna be safe or even work in a man versus a woman. So I think that we have made some really good corrections in this over the past decade or so. I mean, the NIH has put a lot of requirements on making sure that the way that we collect data is diverse, the data itself is diverse. All of these things is something that we've been working on as an industry for quite some time. Which has really led us to this really important inflection point as it relates to AI of making sure that the data that we're training these algorithms on is diverse so that we don't already create the bias. Now, access is a little bit different. You know, when you think about access to medical care, that's already a huge problem in the healthcare industry and one that we recognize as a whole of an industry that we need to improve across the board. I mean, we need a better healthcare system in general. And I think that AI can really help this piece of it, of whether it's being able to use telemedicine and chatbots to help sort of triage patients in rural areas and things that maybe there isn't as many doctors per patient in that area. I think that AI has a really amazing opportunity to actually help better deliver healthcare to groups that perhaps haven't had really good quality healthcare access. As it relates to two-lane medicine and things that are You know, incredibly important to the patients of Louisiana. The access to healthcare is, I think, really important. And I've really, you know, admired Tulane trying to take on AI as a more holistic view of, you know, how do we implement this in our hospital systems to better deliver patient care, which is great. Yeah.

And that's something that our Senior VP of Health Affairs and Dean of the School of Medicine, Dr. Lee Hamm, is really passionate about and is integrating in our hospital partnership. Switching gears slightly, the AIH does more. Like, the AIH is also very focused on information and advocacy as well as bringing its members together. And you recently issued a white paper on the unified patient record. So I have all the questions for you, Elaine. I know you're shocked. So what is a unified patient record? Why doesn't it currently exist? And why is it so important for us to create one?

If you've ever gone to the doctor, And you've had more than one doctor, you've gone to one hospital, or you've moved, you know that there is this huge challenge of having just one record that you don't have to run around to these different places to collect your medical records from different providers. So a unified patient record, it's a comprehensive electronic health record that should consolidate all of the patient's medical history, including clinical data from various sources. So whether you've been to an ER or you have your favorite doctor that you see, your primary care physician, to labs, and even And then the wearable devices that you use to collect day-to-day data of your steps or those types of things, or even glucose monitoring, into taking all of that really diverse data and putting it into one single accessible record. And unfortunately, that doesn't really exist right now. And it's for a lot of really, really complicated reasons. But we have a very fragmented healthcare system. We have incompatible technologies and data silos among providers. So it can be really, really challenging. It's a huge challenge for us. So creating this patient record though is I think really vital for patient care, making patient care more efficient so that you're not having to run around and get a whole bunch, you're putting the burden on the patient to make sure that all of their medical data is up to date and accurate. But it also allows healthcare providers to really make much more informed decisions and hopefully reduce medical errors. So the patient should benefit from this I don't know, bird's-eye view of their health. And then the researchers can actually then leverage all of that massive amount of data to really help provide more personalized treatments. That's a really long way to say it's real, real complicated.

So it's the old Facebook relationship status of it's complicated. Is that where we are?

Yeah, it's complicated. And the AIH, we recently published this white paper that sets the stage for additional papers. So we just kind of went through and ask the question of, like, why doesn't it exist and why is it important? And we're going to have several follow-on papers about this very complicated topic. And it's not just a US problem. This is a problem internationally as well. So this is a problem the whole world is going to have to face.

Because you layer in privacy issues and everything else and the different regulations we have in the US versus Europe, and it gets even more and more complicated than it's simply complicated.

Exactly.

Well, another thing the AIH recently commented on was President Biden's executive order addressing the regulation of AI. Can you explain the principles the president put forward and what the AIH did as a rapid response to it and some of the points the AIH made?

President Biden's executive order on AI tried to cover a number of different principles that ultimately help ensure the public's trust in AI while, you know, making sure that we have innovation and that we have equitable access. And this is across all industries. So this wasn't just for the healthcare industry, but just sort of this, you know, the principles of AI in general. And there's, I think, 8 key principles, and I won't dive into all of those. That's a whole other— each one could be a podcast in and of itself. But there's things like safety and security, making sure that we, as I mentioned, have equitable access and that it doesn't violate civil rights, and then also supporting the American worker. So there's all of these sort of key points. And what the AIH did is just went through each of those principles and talked about what the healthcare industry has been doing, is currently doing, and what we know we still have to do. And Dr. Rafael Rosengarten, who is the CEO of Genialys, one of our member companies, and he's actually one of our board members, sat down and chatted about some of these different problems. And he provided a really robust recommendation on some of these and really also talked about how, as I mean, I keep hitting this point, the healthcare industry industry is doing a lot of this actually really well because we have been using this for many years, and we're also an incredibly regulated industry. So for example, as it relates to safety and security, you know, Biden stated that we have to have these standards to help evaluate and make sure that we're mitigating risk. Well, any AI-based device or invention that requires FDA approval already has to meet you know, a set amount of standards that make sure that it's what we call robust, reliable, repeatable. All of these things that are already in place. So the FDA has already released guidelines on both good machine learning practices and then also AI within software as a medical device. So thinking with like your smartphone, your smartwatches, and the healthcare industry, I have to say, has really enthusiastically— As enthusiastic as we are, really embraced adopting AI responsibly. So the regulatory agencies and policymakers have worked with us in the past to help make sure that we are really making sure that we put up the guidelines for ourselves as an industry, and that these entities like the FDA has really been, I would say, really proactive in making sure that they understand and are able to implement AI into a regulatory process. And then I think the other piece of this that I think is really important is President Biden talked about supporting the American workers because there's this big fear, right, that AI is going to take your job. And I would say that, you know, the way that we utilize AI in the healthcare industry, you know, we really hope to enhance the actual impact of human expertise and that we can have the healthcare, you know, your doctor able to focus on more meaningful tasks and spend time with you in the actual examination room versus having to go through and write notes Or go through your previous records, or as I mentioned, like trying to track down your previous records. If we can utilize AI appropriately and responsibly, it'll actually help enhance the experience of patients in the healthcare environment. But I also think that it will really help drive innovation and provide an opportunity for people that are in the healthcare industry to do more and not just these monotonous tasks.

Before we move on to my final question, I wanted— we just talked about the white paper. We talked about the response to the executive order. Where do people go to find those?

That is actually available on our website, theaaih.org. You can also find us on LinkedIn and follow us on LinkedIn because we post things like the white paper that we have that's also available on our website. That's where you can find all of those things is just go to our website or our LinkedIn page. Great.

And so moving to my last question, you'll be shocked that once again, it's a multi-part question. I know that I like these so much. You got to pay attention. I know the AAIH is an organization-based membership. So my 3-part question is, what types of organizations can join? How can organizations who are members benefit from their membership? And finally, if someone's interested in joining, how can they do so?

I'm going to start with the last one first, which is if you're interested in joining, just talk to me and my contact information, we can provide that in the show notes, I'm sure. But really, you know, the, the AAIH welcomes a diverse range of organizations, including healthcare providers and technology companies. companies that might provide solutions for us, pharmaceutical companies, regulatory bodies, and even research institutes. You know, membership in the AIH, we hope, provides a lot of different benefits, including access to the network of all of the experts and stakeholders that I just mentioned, and also opportunities to really help collaborate on how to use AI in research and drug development, and also policy development, and being able to participate in industry events and having Really access to resources and best practices. And so that we're, we're a really diverse organization, but we also all struggle with the same problems, whether it's regulatory agencies or what to do with large datasets. I mean, these are all things that we all struggle with despite our diversity. So by being involved in the AIH, you can help learn from others that have been doing this and not necessarily recreate the wheel per se. And I would also add, you know, since we're with Tulane here, That Tulane, Rice, Michigan, Stanford, all of these universities have really taken up the banner as it relates to AI. I got involved in AI because of Tulane, and really it's because I want to find ways to help make patient care better at Tulane. And that's really exciting. So, you know, that's another sort of piece I want to hit on. You don't have to utilize AI now to be a member. And in fact, like, you know, my day job is also So I'm in drug development, and my portfolio companies currently don't utilize AI. And I want to find ways that we can and learn from others in the industry that have been doing this. And that is a really exciting opportunity. So you don't have to know AI. If the most you know about AI is ChatGPT, fantastic. Me too. And being a part of this organization, I have learned so much of just, you know, how I can implement AI into my job and my daily life. So So come join us. You have a real opportunity to help build this industry, you know, at a really exciting time.

That is a fantastic closing note. So on that, I want to thank you so much for doing this podcast, which we're dropping at JPM so that we can highlight the AIH and some of the important work in the field of AI, biotech, and healthcare to people at JPM. But I'm hoping you'll be willing to come back and maybe record another one right before BIO and give us another like 6-month update. Do you think we could twist your arm and get you to do that?

Oh yeah, I think you know where I live, so I definitely can do that for sure.

Well, Elaine, thank you so much once again. As always, we will have Elaine's contact info and the informational links on the AAIH and how you can join in our show notes. Make sure you check them out to learn more. And Elaine, thank you once again for joining us.

Thanks for having me. Thanks for joining us for Bio from the Bayou, and we hope you'll join us again.

If you'd like to learn more about the emerging biotech scene in New Orleans and the Gulf Coast region, Visit us at biofromthebayou.com where we have more info on who we are, how to get involved and connected in biotech in New Orleans, and the industry events we'll be hosting where you can meet with us in person.

And we'd be remiss if we didn't give a special thanks to the Accelerator Network for providing funding for this podcast. Learn more about them in our show notes.

We'll catch you on our next episode of Bio from the Bayou.