Welcome everyone to this special event of the Mayo Clinic and Illinois Alliance for Technology-Based Healthcare I’m Professor Neal Cohen from the University of Illinois and I’ll introduce myself more fully in a short while, but first I have the pleasure of introducing Dr. Konstantinos Lazaridis, the Carlson and Nelson Endowed Executive Director of the Center for Individualized Medicine at Mayo Clinic, who will kick off today's event. Dr. Lazaridis Thank you Dr. Cohen for your kind introduction. Thank you to all who have
joined us today for this exciting conference entitled “Technology in Healthcare: A Glimpse into the Future,” with the Mayo Clinic and Illinois Alliance. As the director of the Center for Individualized Medicine, it's my great pleasure to welcome you to this meeting while we enter to the next chapter of this Alliance of our two institutions. I would like to remind you that the Alliance was developed in 2010 to address technology-based health care issues. The Center for Individualized Medicine has been the home and hub of this Alliance, and thus far 25 million in federal funding has been generated by the Alliance. In 2020 alone, 11 research collaborations, 20 translation engagements, and 23 publications took place because of this collaborative effort. Now as researchers we all understand the importance of innovation technology to scientific discovery.
And with this, I would like to describe a quote of an outstanding scientist, Dr. Christian du Duve, who received the Nobel Prize back in 1974 for the discovery of the lysosome. He said, “as scientists we are not simply read the book of nature, we write it with the help of available technology.” how true In 2011, we opened the Center for Individualized Medicine to provide new knowledge, hope, and healing to our patients by connecting patients with practitioners and scientists.
And toward this effort, innovating and sharing technology are catalysts of our new knowledge and critical to our success. Thank you for participating in this conference. Now let's hear from our Illinois colleagues, thank you.
Our thanks to Dr. Lazaridis and to all of you for joining us today. I’m Neal Cohen, director of the Interdisciplinary Health Sciences Institute at the University of Illinois.
Our institute provides human infrastructure that accelerates health research and strengthens collaboration with and among numerous clinical community and campus partners. The partnership we celebrate today, the Mayo Clinic and Illinois Alliance for Technology-Based Healthcare, was established 11 years ago with the goal of translating novel solutions to individualized healthcare challenges. Since then, researchers from the two institutions have collaborated on over 100 projects including patent development and creation of tools and algorithms that are being used in clinical practice today. The Mayo Clinic and Illinois Alliance provides us with a remarkable way to leverage the engineering and computer science strengths at Illinois with the clinical prowess of Mayo Clinic to transform the way patients receive health care. Clinicians at Mayo Clinic collaborate with experts at Illinois in artificial intelligence, high performance computing, and visualization through our educational programs. Mayo
Clinic and Illinois have helped to prepare over 100 students to leverage the power of data and technology and become leaders in individualized health care and health solutions. Illinois faculty, staff, and students, alongside Mayo Clinic physicians, clinicians and researchers work together to visualize the spread of cancer, treat babies with congenital heart defects, delay the onset of age-related cognitive impairment, personalize the treatment of breast cancer, and get faster treatments to patients with depression. You'll hear about more of these shortly. The Alliance will continue to lead the charge to identify global challenges and create solutions. Together we are fostering medical discoveries and innovations, improving student opportunities, creating a path to the future of health care, and making our community stronger.
Now I have the great pleasure of introducing you to my Illinois partners in the Alliance. Grainger College of Engineering Dean, Rashid Bashir, and National Center for Super Computing Applications Director, Bill Gropp, who will talk more about specific collaborations of their organizations within the Alliance. Hello, my name is Rashid Bashir, and I’m proud and humbled to be the dean of the Grainger College of Engineering.
I’m even more proud of our faculty, staff, and students who have worked hand-in-hand with collaborators from Mayo Clinic to develop sensors, devices, and algorithms that change the way a clinical practice is performed. The Grainger College of Engineering's legacy has been established through decades of groundbreaking discovery and innovation, which is made possible through excellence in teaching and research. We are recognized as a top 10 engineering school with now 39 degree programs ranked in the top 10 nationally. Our Healthcare Engineering System Center is working on a daily basis to connect engineers and physicians, with the goal of improving healthcare for all through interdisciplinary research.
Grainger Engineering, among others across the university, helped design and build the Carle Illinois College of Medicine as a first engineering-based medical school on our campus, to produce the next generation of physician innovators. In addition, we have also recently launched two new programs aimed at increasing access to education. The new AI and Medicine certificate will equip healthcare professionals with a foundational understanding of AI applications through real-world medical studies and machine learning. In addition, the new iCAN program improves access to our top-ranked computer science program for non-specialists who seek an understanding of computing fundamentals. I've often said that AI will not replace physicians, but physicians and healthcare providers that don't know AI might be replaced. These programs are examples of how we want to bridge this gap between AI and computing and medicine when it comes to training the next generation of healthcare providers.
Artificial intelligence, machine learning, and deep learning are critical to our future, and one of our major goals as a college is to address grand challenges in healthcare to improve quality, reduce cost, and improve access, meeting this triple aim of healthcare via technology, computing, and engineering, Our partnership with Mayo Clinic places us at the forefront of contributing to this digital transformation of healthcare in order to solve some of medicine's most pressing challenges. We have partners from every department across Grainger Engineering working with Mayo Clinic and we are proud of the work we have done so far, including the creation of tools and algorithms for clinical use. The clinical and patient domain is one that has not been accessible to engineering and computer scientists and systems researchers. Our partnership with
Mayo helps us advance this frontier and has allowed our researchers to fully understand the clinical problem and work on addressing them. For example, we have developed an algorithm that helps clinicians effectively predict whether a patient with depression will respond to a specific type of antidepressant. This reduces the time associated with multiple trials of ineffective antidepressants a patient would typically endure. We've also developed a new technique,e a work that I've personally been involved in, that allows us to reproduce cancerous cell cultures and quickly test many drug treatments against an individual patient's cancer. This significantly reduces the time necessary for physicians to identify the most appropriate therapy to use. This technique is simple, scalable, and suitable for a broad range of applications in drug discovery, regenerative medicine, stem cell research, and biotechnology.
Our students take advantage of opportunities to work with Mayo Clinic through the Summer Undergraduate Research Fellow program. This past year, 11 Illinois students experienced a very stimulating and enriching environment at Mayo in Rochester. Through the Mayo Clinic and Illinois Alliance fellowship for technology-based healthcare, our doctoral students are critical partners in a two-year fellowship, actively translating research into patient care. Some of our best and brightest, such as Arjun Athreya and Yoga Varatharajah, and Dan Wickland, are currently partnering with Mayo doctors and researchers on next generation health technologies.
We are so grateful for the opportunity to partner with Mayo Clinic and I very much look forward to our continued collaborative efforts to make our relationship deeper and stronger and to continue to expand our impact on clinical practice and patient care. Thank you. The National Center for Supercomputing Applications is a leader in advanced computing software, data networking, and visualization resources. Since 1986, we've been at the epicenter of supercomputing research, pioneering innovations in technology and using them to solve the pressing questions of the day. From the first popular graphical web browser to groundbreaking research in medicine and astrophysics, we don't just push the envelope. At NCSA, we give it our own unique stamp, taking Illinois innovation into the forefront of research worldwide.
The human body is infinitely complex, with thousands of genes that dictate everything from the color of our eyes to our tendency to develop cancer, at NCSA, we help researchers decode its mysteries. Leveraging our expertise in data analytics, software development, simulation, cyber-security, and visualization drives innovations that could lead to longer and healthier lives, and we're only just beginning when forging critical relationships with partners such as the Mayo Clinic. With NCSA at the helm of computing and technology, our collaboration with Mayo enables us to build visualization and analysis software tools that help researchers study data and present complex data to clinicians to interpret results efficiently. Our solutions and advances have been as unique as the problems we face, such as developing web-based tools to process and visualize microbiome data allowing researchers to study how a person's gut might contribute to disease, or creating new tools to improve current computational methods to accelerate the genomic processing pipeline and ultimately reduce the time to process a patient's testing from 20 days to a little over half a day.
All this has profound and positive impact on patient care. And there is a continued need for the future of technology-based healthcare, with an increased demand for improved computation combining more data types, more biological and tracking data, and image data. NCSA continues to provide HPC resources for current research needs. At NCSA, our experts work with our partners to develop innovative solutions that bring the power of advanced computing to bear on the most challenging problems. Our focus is on
that last word in our name, creating applications that are transforming our world. Along with Mayo, NCSA is driven to contribute to the interdisciplinary effort needed to understand human health better. NCSA provides a diverse, supportive environment that enables innovation today, tomorrow, and for years to come. Thank you. With thanks to our partners, the Grainger College of Engineering and the National Center for Supercomputing Applications, and on behalf of the University of Illinois, we look forward to a very bright future for the Alliance indeed.
I am now delighted to introduce you to Dr. Andrew Limper, the Walter and Leonore Annenberg Professor of Pulmonary Medicine and professor of Biochemistry and Molecular Biology at Mayo Clinic. Dr. Limper is a consultant with the Division of Pulmonary and Critical Care Medicine at Mayo Clinic, and also a proud two-time alum of the University of Illinois. Dr. Limper was not able to join us today, but he has kindly shared remarks with us via video.
Hi. It is my pleasure to add my warmest welcomes to this Mayo Clinic University of Illinois Alliance meeting that is focused on technology and health care. My name is Dr. Andy Limper. I’m the Annenberg Professor of Pulmonary Medicine at Mayo Clinic, and over the last four and a half years, I've had the privilege to work in the Kern Center for the Science of Health Care delivery at Mayo. When I completed my undergraduate work and medical school training at the University of Illinois in 1984, I entered into a very different world of health care as a professional. It was a really quite a different world at that time.
Every single bit of medical record was captured by paper, x-rays were all maintained on these plastic celluloid hard copy films, and they had this very nasty habit of becoming lost when least convenient, like when you were taking a patient to the operating room. In addition, our understanding of a patient's history came by pouring over stacks and stacks and stacks of these records and films in the wee hours of the morning, and in 1984 that was the job of the of the intern, me. Fast forward 35 years or so, what a bold new world we're living in now. Over 95% of health care records in the United States are maintained in some type of electronic format. And surprise, surprise, the largest electronic health record vendors actually talk to each other, that's a great step forward. In addition, most all of the x-ray images, the echocardiograms, catheterizations of the heart, the brain, the kidney, they're now all in digital format.
We also have a whole myriad of electronic tools that give us risk scores, so we know when there's going to be a severe outcome, such as when a patient is at risk of being re-hospitalized, when somebody looks like they're developing sepsis, or a clot in the lung known as a pulmonary embolism, and a whole host of other problems, not to mention all the other technologies in molecular genetics and molecular biology, biochemistry, engineering, and all the other sciences. And probably the best evidence of this is, because of all this technology, mankind was able to develop new vaccines for the worldwide plague of COVID in a record amount of time. And if you watch the news on any given night, there are new drugs being released every week for a whole myriad of chronic and acute medical illnesses. But I don't want to paint the picture that things are perfect or ideal.For instance, not all the electronic health record systems actually interface with each other, and it's actually a challenge to actually look at films, despite the fact they're captured in digital format across these different electronic health records systems.
Now I've been a practicing lung physician for about 30 years. I don't really care so much what the local radiologist has to say about what the CAT scan looks like or the chest x-ray looks like, that's my job to look at the actual films, and sometimes, oftentimes, I still have to have some of those pushed to me electronically, which takes time, or worse yet, have to come by snail mail on a disk. So there's a lot to do to improve the electronic records. I also mentioned we have all of these risk calculators. In
fact, we have over a couple 100 of them or more within the major electronic platforms, but there's still studies and data that indicate that providers don't use them on a routine basis. Why? They don't fit into the busy clinical workflow or they make you go over here over there to find those calculated values when you are already within your electronic workspace. We need to teach these algorithms to work in the electronic background to identify patients at risk and then raise the awareness, raise the awareness to the provider in real time when that information is most needed. I've actually likened it to have that, when I was, in 1984, that electronic intern that knows all all the information on the patient and can raise the information to the attending advanced practice provider when most most needed. We have a ton of information but we have to tackle that electronic behemoth to make our patients’ lives better.
We can actually learn a lot more from the information that's already there, there's all sorts of waveform data. So for instance, up till a few years ago we've collected thousands of electrocardiograms looking for rhythm problems or myocardial infarction and now we know that there's subtle electronic signals within the ECG that can actually predict the ejection fraction of a patient, that's how well the left heart is squeezing. A lot of important information that you really can't see with the human eye but machine learning can train that. Simple blood tests and complex blood tests can help produce the risk of malignancy or other chronic diseases. So, where we're at is we need to use these tools, technological tools, electronic tools, molecular and other tools, to actually prevent disease, or at least diagnose it at its earliest possible moment, actually before the patient actually knows they are unwell, and that allows earlier diagnosis and allows the patient to get back to normal. So tomorrow is going to be bold, tomorrow is going to be bright, we need talented young minds and old minds to harness this technology, to train the mountains of electronic data, to get all of the information out of the records the images all those electronic waveforms, and bring all these new new tools together to make the lives of our patients better.
Lastly, we also need to use technology to democratize health care. There is a great chasm in healthcare equity in the United States, emphasized during the COVID pandemic, but also worldwide. It is the challenge of us to use this technology to make health care equitable to all and available to all, so these are our challenges. These are your challenges:
enjoy the series of lectures in the conference that is going to point us towards the future. There's actually more than plenty of work for all of us to do. Thank you and enjoy. I now have the great pleasure of introducing Dr. Claudia Lucchinetti, who is chair of the Department of Neurology and Eugene and Marcia Applebaum Professor of Neuroscience and dean for Clinical and Translational Science at Mayo Clinic Alix School of Medicine and director of the Mayo Clinic Center for Clinical and Translational Science, and finally, who is live rather than pre-recorded. Hopefully Zoom will treat
her well right now. Dr. Lucchinetti. Well thank you for the opportunity to speak with you all today, and my talk today is going to be on advancing clinical and translational science and digital health, and it's really an opportunity that I welcome the chance to talk a bit more about the next chapter of the Mayo Clinic and Illinois Alliance. So in terms of digital health, we know that COVID really did accelerate the need for digital health by providing tools that can reach and facilitate care remotely, but really, it's the importance of how do we use digital health to drive technology that actually improves health and wellness. The application of digital health through disruptive technologies and cultural change is going to have a transformational impact on the practice of medicine and every consumer, consumer of healthcare. And also, digital innovations are transforming translational research.
It's important to recognize what is digital health. It's everything from wearable gadgets to ingestible sensors, from mobile health apps to artificial intelligence and virtual reality, from remote patient monitoring to telehealth, from robotic care to electronic medical records in deep learning algorithms, and also the use of decentralized clinical trials. But the promise of digital health includes the possibility of preventing disease, monitoring and managing chronic conditions, hopefully shortening disease duration, easing symptoms, tailoring medicine to a patient's needs, diagnosing new diseases, anticipating worsening illness, improving the quality of life. In addition, the goals and promises of digital health are to reduce the length of hospitalization, reduce the care of health care, the total cost of health care for a patient, and make health care, in general, more affordable.
In addition, the promise is to collect and integrate data from a variety of sources to enable real-world research, and to change how clinical trials are conducted to make patients the point of care with new roles for clinicians, patients, and caregivers. But there are challenges to realizing the full potential digital health. It must first and foremost be shown to actually improve the medical practice and human health. There has been an explosion of AI publications and healthcare, but the reality is that few have really been adopted to show a change in care of patients. Clinical trials of AI interventions on meaningful patient outcomes are actually few and far between.
We need rigorous approaches to develop and test innovations. These need to be developed and standardized. And of course we need to recognize that digital tools can have unintended ethical consequences, including exacerbating health disparities. It is an absolute priority when we talk of engaging our two centers, but we also think about including rural residents, racial and ethnic minority groups, and other end users, in the design of technologies to enhance adoption and their use. And even when tools are proven, translation to real-world application requires navigation of regulatory hurdles and application of dissemination implementation science principles, translational research is critical.
The overreaching priority really is to facilitate the creation, validation, dissemination, and implementation of innovative digital approaches to health care and translational science. And to do so, we need to foster translational research that advances innovations and digital approaches across the healthcare continuum, and we must ensure that issues of health equity are fully considered in the translation of novel digital technologies from discovery to application. This slide here shows the approach at least from the Center for Clinical and Translational Science, where I’m a director, to catalyzing digital innovations. The key areas that will enable digital technologies are data, development of algorithms and technologies, testing of digital tools, and dissemination implementation of digital tools. Equity around digital innovations will be emphasized and specific components, such as compliance, ethics, and regulatory issues need to be supported throughout their development.
And here you see the importance of where data science and informatics is so critical in the generation of the models and the importance of systems engineering and informational technology to really drive from a model, where we really want to ultimately see it scaled and disseminated throughout, with the dissemination of this model beyond our local environments. Now the development of algorithms and technologies, there's been considerable interest. Edge capabilities, such as custom wearable and implantable reference platforms, can be developed, that allow for data acquisition from novel combinations of sensors with the capability of executing machine learning algorithms in real time. But this requires calibration and validation, especially for testing in live environments, which tend not to perform as well in the real world, while also the imperative of creating meaningful standards. We need to connect investigators with mobile app developers to enable development and testing of tools.
And a key challenge for these tools, I said, is their performance in real life settings. Many are developed with existing or retrospective data and also validated on existing data. We need to ensure that digital tools that provide diagnosis, treatment recommendations, or testing recommendations that lead to improved health outcomes and are safe for patients and consumers. Prospective approaches are needed such as patient-level randomized trials, pragmatic or clustered randomized trials, or observational studies. And this is a real-world example of the development and testing of an algorithm to identify patients with low ejection fraction using electrocardiogram, the so-called “EAGLE study.” Step one was
accessing and creating data on EKGs, echocardiogram, and EHRs, in partnership with our informatics core. Step two: developing and validating the algorithms to protect low EF in the general population and assessing its performance in such key subpopulations. Step three: to partner with informatics to implement the algorithm in the silent mode in the live EHR to ensure algorithm performs as expected. Step four: prospectively evaluate the impact of the algorithm on patient outcomes, In this case they conducted a pragmatic randomized trial, the EAGLE trial to provide real-world data on the algorithm's ability to detect low EF in primary care systems by partnering with community physicians across the Mayo Clinic health system at the point of care. And finally, to implement the algorithm in routine clinical practice and identify opportunities to scale the algorithm beyond the Mayo Clinic. Promoting equity fairness and reducing bias is fundamental at all stages of the development of digital innovations.
Since they can all be affected by or result in bias or inequity, we must understand the potential biases associated with our data source, we must recognize that the recording of data on our sources such as EHR are often incomplete or inaccurate with respect to race and ethnicity, we must recognize that our native analytical decisions can perpetuate racism and inequities and they risk baking inequity in the system by interpreting racial differences in the underlying data as immutable biological facts, rather than as reflecting the societal effects of racism. Without mechanisms to monitor and understand poor model translation and performance in the life setting, real world impacts that could harm racial and ethnic minorities might be missed. So we must develop these tools and practices to improve generalizability, documentation quality, transparency, and reproducibility for ethical and race sensitive data driven insights. And these barriers - what are some of the barriers and gaps? We need cloud data management which is PHI secure. We need robust software development and a clear strategy, including interoperability, modularity, reusability, and scalability. We need quality management systems and regulatory support, our researchers can't be expected to figure that all out. We
need a compendium of validated digital endpoints developed by applying algorithms to wearable data, as well as digital algorithms from external sources across disease areas and sensors and make it available for investigators. We need pathways internal and external, for the development of devices to support novel algorithms and sensors. And then we have to integrate all of this in the current delivery system, so that as algorithms evolve and advance as they are built to purpose within the healthcare system itself they need to also facilitate new insights and continuous grounding and clinical validation every step of the process, a virtual circle and continue, that continues. So a decade of success in collaboration, we can speak about the Mayo Clinic in Illinois Alliance, and we need to recognize some great examples. We've heard of Dr. Arjun Athreya, and you'll hear more from him, I call it a case study and the power of what's possible. With his training beginning at
UIUC, traveling through his PhD, now a Mayo faculty leading tremendous work from theory to practice, leveraging engineering methods for individualized medicine. We heard about his important work in the point of care and psychiatry to enable genomic medicine, showing how clinical data together with a validation cohort, together with the right team approach, really can lead to significant impact in predicting gene panels that support the response to various antidepressants. And then we have the tremendous work of Dr. Worrell and team students, such as Yoga and Saboo and others, have really joined from UIUC, advancing tremendous work in brain health, with scalp EEG, in the classification of epileptic focus with investigative EEG, on the classification memory networks, human memory and mood prediction, and seizure forecasting.
So my last two slides are really then what are the foundational principles we must rely on to continue to enhance and build on these and other successes from the Alliance. One, it's essential that teams are working together, that we foster this to solve actual problems that matter to health care and human health. It's important to get a grant, it's important to get a, a measure that has a return on investment, however, ultimately, it's the impact on human health that is foundational to the types of problems we want to tackle. We need an inclusive approach, we need developers of AI and digital tools that work together with healthcare workers and other stakeholders throughout the design and creation process, through to deployment.
We need boots on the ground, these were great examples of teams working side by side. Zoom and slide decks will not be enough in this transformative time of innovation. We need to enhance our partnerships with computer engineering, so robust and excellent with the UIUC team, trained to build products and solutions. We need opportunities for support with the entire spectrum of data management processes and operations. Another
tremendous strength of UIUC in collaboration with Mayo. We need to cross- fertilize synergizing across complementary expertise, and we need to develop and understand what the clinician and patient are dealing with, developing a common language, and ultimately building a shared education framework. So just to close here, this is a slide that Neal gave but I really liked it because it highlighted just at the center, so many areas of intersection on AI, on computational genomics, multi-omic pipelines, sensor divisor work, ML and base clinical discovery, visualization and image analysis. And at Mayo we can really underscore the clinical needs, bring that application flavor, highlight our needs within IT, demonstrate and talk about and address our collaborative challenges, seeking to integrate and translate this work to have a meaningful impact on human health, and partnering closely with assembling interdisciplinary teams with UIUC, which will incorporate academic research and development, leveraging technology-based clinical solutions, and again, integrating and translating with purpose.
So with that, I will close my presentation and really thank you for your attention and the opportunity to speak at this conference. Thank you so much Dr. Lucchinetti, really wonderful. We now move on to our live panel discussion on success in education research and translation of technology-based healthcare, featuring some key Mayo Clinic and University of Illinois collaborators, alumni, faculty, and staff. Allow me to start this section by introducing Professor Ravishankar Iyer, the George and Ann Fisher Distinguished Professor of Engineering in the Grainger College of Engineering at the University of Illinois. Ravi, you get the first question. [laughs]
You've been a mentor, [laughs] a mentor, a PI, and a collaborator on many many projects during your career and a key player in the Alliance. Can you describe for us the importance of Alliance collaborations in education research and translation? Who better to do that than you? Thank you, thank you, thank you Neal. Actually, I should thank Dr. Lucchinetti
for giving such a nice talk, because it had all the essence of my answers, and if you listen carefully, I think she described effectively how important it is that that we educate ourselves, we as engineers effectively educate ourselves, when we go to Mayo and work with our are really, totally brilliant Mayo Clinicians and their, and their researchers, we come up with the best ideas and, and indeed many of these lead to translations. I think the best way I can describe it is to use a couple of these guys who, who Claudia, Claudia described in her slides. You know, Arjun was a quintessential risk taker and when I, when, when Professors Winterbaum and Wong, suggested that we sent somebody to Mayo. I was thinking, “what would this person do?” and I approached Arjun and said, “Arjun, will you go and being a risk taker and, and, uh, you know the kind of, uh approach to research he had, he didn't hesitate for a moment and said, “yes,” and, and I was a little bit worried, but once he got there, whatever, I mean, Dick had said to me that, “If we, the best way for us to put people together and, and to, to do great research is to send our students there. Once they are embedded among our interns and, and, and postdocs, as Dick said, “Magic will happen,” and truly, magic happened, and in this case, it was a project to look at drug efficacy in, in patients with depression.
You know, it was easier to take the next, next risk uh when, when [inaudible] and yoga and others went um, but I think the the real magic is that unlike many other hospitals that I have worked with, at Mayo you have people who are passionate about their patients and passionate about research, and you see it the moment you walk into their offices and start talking to them. And this more or less takes our best students and energizes them that they can really do something to help people at large, and this combination between our excellent computer scientists, engineers, neuroscientists, what, what have you, and the totally number one clinic in the world at Mayo, I think I believe produces the magic that I haven't seen elsewhere, and truly you know, we work with physicians at all the top hospitals in the world, true, and, and I'd say that the Alliance is unique, we made the magic work, we made the research work, we made the education work, and what's important is, we not only have invented new patents, but we also, and translations, we're just starting to help patients, it takes a long time before it gets into the hands of the patients or into the hands of the physicians who are really dealing with the patients, and that has also happened. So let me, let me uh stop there and then turn to, to Nick and say Nick let me, can I maybe ask you my first question. I think you completed your postdoc at Illinois, but I also know that in that process you're influenced by a whole bunch of people, some outstanding folks, not just from engineering and physics, but also you know from, from crop sciences, from agriculture. How did that experience really help you or prepare you for your role that is, that, that's emerged to be, you know, in, in translation research, and your leadership that you now have in in that field at Mayo? That’s a great question, and uh, I just, I’ll say, Ravi, I think one of the things we under appreciate is that, I think, great translational and and innovative, uh, production, comes not from necessarily following the straight-line path but really being able to attach a lot of insight across different teams and learn from different members in a diverse community of academics, so you're right, I worked with animal sciences, I worked with people who are focused on, on agriculture, I worked with Bryan White, I worked with Carl Woese, um and one of the problems I was focused on for five years was the origin of life, which taught me so much about how to understand information from, from traces, right, how do you infer more from, from not being able to measure everything, and that really, that data focus, right, because it's not like we're going to do a lot of experiments in the origin of life, we're not going to try to create life too many times, too many ways, it really forces you to come up with new approaches to data, that can be applied to medicine, it also teaches you how to work in a very interdisciplinary environment, right, and team medicine is something so central to Mayo, and team science, I think, goes hand in hand with that, so I think the the environment at University of Illinois in some ways is really ideal for, um, bringing, bringing, a broader vision and spark into, of academia, into and mixing it with that broad vision and team, team medicine approach at Mayo, and it's, it's really been a wonderful journey.
Just a quick, quick question, I can't let you go without this, it's uh, “what's on the horizon? I really want to know.” So I’m biased and you're probably biased [Laughter] so we, we probably share the same bias, but I, I think, you know, data leadership is absolutely one of the things that will define the next 10 years of many industries, not just medicine, and I think, I think having that data leadership is where that next step in the horizon is, and, and we heard Claudia talk about that and, and what's needed for that, and the vision for that in the future, but I think if we don't wrestle that down to the ground from within medicine, we're seeing all of these AI and technology-based companies try to essentially wrestle the, the problem of medicine down from the outside, and I think that's a mistake. So this is something that we have to invest in, and I think if we don't do it, we'll be worse off as a society if we don't have that vision coming from a place like Mayo.
Thank you, thank you very much, Nick. Let me, let me turn to you, Arjun, uh you know, of course, you and I know that you're one of the early, early Alliance Mayo Illinois Alliance fellows, how did that really help you or, or led to where you are today, and you know, can you let it be a little bit more specific, uh, about your work from, uh, research to clinical practice, you know, that's a, that's the experience you uniquely have as a student. Hi Ravi, and thanks for including me on this panel, um, I think to me the biggest, uh, perks of the uh, the fellowship, was the travel part, that money that Mayo gave us, right, we were given, I think, five or six thousand dollars a year to come and spend time in Mayo, and, and what it really allowed us to do is, uh you know, come in and understand, you know, the, the, the first needs of the practice that can eventually meet the needs of the patient, uh, you know, and as, uh, Dr. Lucchinetti rightly said, uh, or even as Dr. Limper also said, you know, it's important that we build technologies that make sense and can be used and can be implemented and transform the practice, and so it's not possible where people can just throw data sets over the internet and say, “now give me a p-value value and a result,” and then we'll come back and build an AI model that leads to no viable outcome, instead, I think the fellowship forced that shoulder to shoulder work where clinicians and I spend copious hours and weekends and holidays together, working in these conference rooms, you know, writing equations on one board, clinical significances on the other glass board, and eventually match that, that, you know, get that integration to the point where we could design clinical trials that were, um, AI, you know, based and, and then eventually, you know, start monitoring its efficacy in the real world. And I think that formula, what
began with depression has quickly moved beyond depression, at least as far as I’m concerned, in my day-to-day life at Mayo, uh, we are now looking at pediatric behavior, we are looking at, uh, surgical outcomes, uh, you know, in those with brain hemorrhage, and we're also looking at other diseases that were certainly not part of initial conversations, but now that formula is becoming much, much more easier for everybody to understand and appreciate and, and move, um, more quickly beyond that. So I think that's really what the Alliance did, and it, it spawned a career for me, right, and so I think without that I would not have actually understood the body language of the clinicians and really see why, why is it that they could work with an engineer, um, so I think, to me, that was the biggest perk of having the fellowship, and, um, I really, I mean to all of us it sounds like the travel part of money is small, but believe me, that's the only thing that made it possible for us to have boots on the ground. Thank you, thank you very much, Arjun, and thank you for both you and Nick for your contributions, and Nick, Neal, back to you. Ah thanks, thanks folks, um this question is to Colleen Bushell, who's associate director of Healthcare Innovation at NCSA at the University of Illinois. Colleen, you've been instrumental with developing the Alliance since the beginning providing ways for NCSA and the University of Illinois to work together with Mayo Clinic physicians to understand computational approaches to supporting health care, much like we've just heard, and by the way, so people here know, you're the one who provided that fabulous slide that Dr. Lucchinetti used to
close out, close out her wonderful presentation, so credit where it's due. Um, my question, “can you highlight us please, if you would, highlight a specific project that helped to translate research into clinical practice?” There's nothing like hearing really concrete examples. We've heard several so far. Please, please take this moment, if you would, to, to highlight one you, you care about, and more generally, how do you facilitate collaboration? How's that an example of the facilitation and collaboration? Great, thank you, Neal, for the opportunity. Um, it's hard, over the last 10 years, to focus on just one example, there have been many exciting things, so I did just want to comment on a few early engagements that really led to the way we approach projects now. And so one was working many years with Bryan White here at the University of Illinois and several different physicians at Mayo to really work on identifying biomarkers, working on data analytics, and then also working with Matt Ferber, who's at Mayo clinic, on the design of, uh, patient reports from genetic results of a colorectal cancer, but, and I think both of those really pointed to many of the issues that now we really work on and focus on in the translation area. So, so a project
right now that we're working on, that I would say is a result of some of the concepts that we worked on with Matt Ferber, is a project we're working on with Dr. Freimuth right now at Mayo Clinic, and the visual analytics team here at the University of Illinois, on designing, um, some visual representations of results from complex models, predictive models that are then used, can be used in the, “Epic,” which is the medical record system, so that clinicians then can find the actionable information and, and really discern what they want to do with that, that information. So we're working right now on, uh, specifically, uh derivatives of the algorithm that you just heard talked about today. And the other one that's,
that I think is really valuable right now, is a current effort with, uh, Dr. Charles Blatti, here at the University of Illinois, along with Mohammad El-Kebir, who's a computer science faculty member, and Nick Chia, who you just heard speak from Mayo Clinic, in their work in tumor evolution, and so, I think a lot of the work we've done in the past with analytics showed that if we could start to encapsulate some of these powerful approaches into a reusable software tool that it makes it easy to execute the computation, uh, includes visualization to understand the results, that that will really help to, um, to spread the development of that research, and so what we're trying to do now is, as we're working on this tool as a research tool, at the same time, we're thinking of it as eventually a tool in the clinical setting. So we're working with clinicians on how they would use that information, what do they need to know to make decisions from that information, so that we can eventually, um, get to that point that it's translated into the clinical environment. So I really think looking to the future, um, advances are happening faster than adoption, and so the area of healthcare delivery, um, really trying to tackle some of that is, is very critical. Wonderful, love that vision of the future, thank you.
Question now for Eric Klee, who's director of bioinformatics at Mayo Clinic’s Center for Individualized Medicine. Eric, you've also been involved with the Alliance, uh, for a long time and in many different ways, reflecting your many leadership roles at Mayo Clinic, and easy for us to see that your engagement continues to evolve and grow, much to our joint benefit. Um, what's the role, from your point of view, what's the role of biomedical informatics to individualized medicine, and how do you see that evolving in the future, so your version of the future. Neal, that's a great question, and actually, it's a, it's a loaded question, we could talk for hours on, but I’ll try to be succinct here. If we think about individualized medicine, it really has many different meanings, depending on the context in which you look at it, but, if, over the last 10 years, at the heart of it has been the molecular characterization of patients and how that can inform their health care, so when we think about bioinformatics and the impact that it has had on individualized medicine, you know, it's gone through an evolution, where very early on, a lot of the work was around how do we simply get information out of the massive amount of data that was being generated by these sequencing instruments. You know, in 10
years, we've had tremendous growth and explosion and innovation around sequencing technology that's allowed us to profile patients at a greater and greater level, and a lot of the work that was done early on was around generating efficient, accurate, uh, workflows that allowed us to generate information, but we saw a shift of that over the last few years, where we had now moved into more interpretive applications, and understanding really, how do we get to the [inaudible] of individualized medicine and using that data to inform health care decisions, and there we saw investments in interpretive algorithms and integration of data to drive things like pharmacogenomics, and cancer profiling for treatment decisions, as well as diagnostics in the setting of rare Mendelian diseases, and these are very early wins in which we saw bioinformatics teaming up with clinical practice and genomics to really meet the division of individualized medicine as we kind of set out 10 years ago. But if we look at where we're going, I think there's again a shift that's taking place, we're moving more into an era of connectivity and integration, sequencing is now so pervasive it's a commodity, we talk about sequencing now in studies of the size of hundreds of thousands of individuals or patients, or millions, population level sequencing, and yet we still want to focus back on the individual, the N=1, and what it means to their health care. To do that, we really have to build the systems that are capable of integrating this data and to think about how do you interpret the results of an individual in the context of the knowledge that we now have worldwide across entire populations and what does that mean, and not only in genomics, it's now very multi-modal, and I think that's, that, that's the key piece of the connectivity that is emerging right now and why things like AI and data science have become so prolific, is that we're creating structured data sets that allow us to bring together genomics, clinical data, imaging, digital pathology, wearable data, and others all to profile the characteristics of the patient. So now you need to think about that moving into the integrative phase of how do you build intelligent systems that can integrate that information in an almost near real-time way, combining in the emerging knowledge in the field, because what we've seen in the last five years is just an explosive exponential growth in our understanding of genetic disease, and that's changing on a weekly basis, and so we need to continue to integrate that into the decision making. And then probably the biggest key to this, in the next three to five years, is how do we take all of that connectivity and integration and then present it back to our clinical experts and our patients in a way that's intuitive for them to understand We cannot expect them to become genetic experts, you know, we're not going to have clinics full of molecular geneticists, so we're going to have physicians focus on what they're doing and how do we build the tools and present the information in a way in which those individuals can routinely integrate it into their health care activities and inform patient recommendations without even thinking about the fact that that patient's genome is sitting on the back end of these decision-making processes.
Wow. That's fantastic. We're getting close to the top of the hour and we have a hard stop at one, so perhaps I could circle back around to um Dr. Lucchinetti or Dr. Lazaridis, if you'd like to add some final integrative kinds of remarks. I
mean, the themes you folks started with are the ones that our speakers here in the panel have provided such nice examples for and, and gave a glimpse of the future where we would would like to go, um, if you'd like to take another quick comment, [overlapping conversation] When we speak about boots on the ground, I think it's in both directions, um, I think we, having more opportunities also to bring our clinicians and those who are really driving in this space, into the UIUC environment as well, I think that's going to be important, um, but I think overall there, those are great examples, we want to be able to scale that and accelerate, um, and support that type of work, so just looking forward to, to understanding what's in the pipeline and different mechanisms to support that work. You know, Arjun talks about needing five thousand dollars. I mean seriously, like I thought we should, we need to understand who is there and ready to go and have opportunities, because there's no shortage of ideas and we just want to be sure that we direct and help those ideas move toward a direction of shared purpose and goals around really having a meaningful impact on our patients, on our society, on our health care process. Kostos, I’ll head over to you to close this out. Yeah, thank you, thank you, Claudia, thank you, Neal, uh, this was a great interaction, a great meeting, were also fantastic ideas, and we're all excited about what the future holds. In my early paradigm about [inaudible] I, I would like to make the comment that he knew about there is an organ, uh, within the cell that he called the “lysosome,” but he couldn't prove it until, uh, electromicroscopy became available, and this is the tool with technology and so us practitioners, you scientists, all of us, we know that the, the solution for our patients but we have to have the technology and the proper minds put together to make these discoveries, so technology is going to be profoundly important for the future. At the same
time, we have to go back to our primary value, which is the patient, and to make the patient the center of this effort. So as we move forward and, and we advance the technology the ways that we do it these days, we have to still remember that we do this for the patient, this is the ultimate goal. And so with this, I hope the series will continue and we have more leads to new discoveries to improve the care of our patients and [inaudible] the community. Thank you. Thank you, um allow me to say that every time we get together it is exciting, we always come away thinking we don't spend nearly enough time together, so thanks for all of you who participated, thanks to the fabulous administrative support by Deborah Miller, Lisa Colburn, Nicki Nicholas, LeaAnn Carson, and Maria Maragos, to the NCSA tech crew for video and streaming capabilities, and to everyone in the audience for showing up today, it was really a pleasure.
2023-09-19