 I'm very happy and proud to be able to introduce to you Professor Batali Rice, an outstanding member of our faculty in every regard. He received his PhD from UC Berkeley, and then he did a very interesting postdoctoral experience at UC San Francisco Medical School, which truly brought him into the translational clinical domain. And after a brief stint as an assistant professor at the University of Wisconsin in Milwaukee, he joined our faculty and obviously was recently promoted. His expertise is computational fluid mechanics, but done in a way where he strives to have his model predictions clinically relevant. And I literally mean to a patient where he takes imaging data from patients and matches them up in very clever ways with his computational models and is able to give clinicians insights into what therapies are needed in order to help patients, for example, with cerebral aneurysms, which are very, very difficult surgeries. And Batali does a tremendous job in linking, in a sense, what we might call hardcore engineering with clinical medicine. He's also an outstanding teacher with us, very, very dedicated, particularly at the undergraduate level. He's completely transformed our biomedical transport course. And he's won our Tacker Prize for the outstanding teacher in our school. So without further ado, I'll turn it over to Batali. And again, a delight to have on our faculty. George, thank you very much for the kind introduction. I hope you can see my slides. Let me make sure you can see my slides here. Yes. And it's my pleasure and my honor to give this talk. I will talk about image-based modeling of blood flow. And throughout my career as a biomedical engineer, I see myself as the bridge between engineering and clinical communities. And that started back in my graduate and postdoctoral studies at University of California Berkeley Mechanical Engineering and Radiology Department of UC San Francisco. And I was driving back and forth between these departments across the Bay Bridge, Oakland Bay Bridge that you see on this slide. And I'm very happy that now at Weldon School of Biomedical Engineering, I can work with outstanding engineering students and colleagues and keep my close collaboration with clinicians without physically driving across the bridge twice a day. So my cardiovascular flow modeling lab is working on improving diagnostics and treatment of cardiovascular disease by conducting engineering analysis of blood flow dynamics. And due to the limitation of time, I will talk only about those of my projects that are funded by National Institute of Health. So I will briefly talk about modeling blood flow in brain aneurysms, designing a filtering device for removing toxins from blood and image-based modeling of cerebral spinal fluid and brain health. And I will talk about my clinical collaborators, how I got to work with the leading neurosurgeons, radiologists, and medical physicists in the country at several clinical centers. So historically, the main focus of my work was on image-based modeling of blood flow in brain aneurysms. And those are local dilations of arteries providing flow to the brain. So if they grow and rupture, they cause hemorrhage and stroke. So the idea is that by looking at blood flow dynamics, we can help with pre-certification identifying any rhythm that are likely to rupture and providing guidance for surgical treatment of these patients. And that started with my post-doctoral studies at Radiology UC San Francisco with David Saloner, who is a professor of radiology and director of Vascular Imaging Research Center. And we looked at monitoring aneurysm patients with annual MR and geography data. And you can see registration of images obtained from clinical studies, MR studies over years. And we found that out of 100 patients that we studied, only about 12 aneurysms actually grew. So most of the aneurysms remained stable. And I started modeling blood flow dynamics in these patients to understand how flow descriptors are linked. I'm trying to advance my slides. How the flow descriptors are linked to aneurysm progression. And we found that indeed, aneurysm that grow tends to have low wall shear stress, frictional force between the flowing blood and vessel wall. And so I applied and reviewed my first translational NIH K-25 award, K-25 award that helped me to start my work as independent investigator. At that time, I started working with Dr. Michael Lowden, who is a world famous neurosurgeon treating these patients. And later on, he moved to become president and CEO of Barrow Neurological Institute in Phoenix, Arizona. So I started modeling these surgeries with a computation fluid dynamics to predict postoperative flow fields. And we developed this virtual contrast method of visualizing blood flow and indicating regions with increased flow response time that could predict blood clot formation in aneurysm following surgeries. So we were supported by NIH Heart, Lungs and Blood Institute, larger R01 grant and successfully predicted postoperative clotting in a number of these patients. So at that time, I realized that to continue with my studies, I really need engineering students and I need to be in the engineering department. So first I found this position in University of Wisconsin-Milwaukee with also appointment in medical college of Wisconsin and University of Arkansas for medical science. And so I modeled a number of their patients with our predictive models. But fortunately, I was able to move to Purdue and join this wonderful department here where I can work with the best students in engineering and maintain my collaboration with clinicians. So also when I moved to Midwest, I started working with Michael Markle, who is a professor of radiology at North Western University. And he is a pioneer in developing MRI methods for measuring blood flow in vivo. He coined the word 4D flow MRI, time-resolved three-directional velocity measurements. So by cleverly manipulating magnetic field gradients, you can get velocity vectors over the cardiac cycle. And comparing these images of blood flow with MRI to our computational models, we understood that each technique has advantages and limitations, modeling assumptions of computations, limited resolution and noise of MRI. So the question is, can we somehow fuse this data? Can we blend these two techniques to overcome limitations of each approach? And since we don't know the true flow in a patient, we need some benchmark data, some test data. And that's how I expanded into experimental studies. And we developed a method for 3D printing transparent flow phantoms for experimental flow measurements from medical images. And I started my close collaboration with Pavlos Vlasov in beginning engineering here. And in his lab, they use particle image velocity symmetry to obtain highly resolved, time-dependent volumetric data in these phantoms. In parallel, I started working on in-vitro and in vivo 4D flow MRI measurements at Purdue. And you see measurements in flow phantoms and in healthy volunteers obtained on 3 Tesla scanner at Purdue MRI facility. So this work is supported by NIH R21 exploratory grant between the team of Purdue and North Western. And we are developing methods to enhance resolution of MRI so that we can better get relevant flow descriptors. But ultimately, our goal is to identify aneurysms that are at risk of growth and rupture. And so we want to combine clinical risk factors that are currently used in clinical setting with our engineering analysis with the risk factors related to blood flow. So for developing predictive models, we need a lot of data, a lot of patients. And my connections to clinicians at several centers helped to put together a team of investigators from Purdue, North Western, Barrow Neurological Institute, and UC San Francisco. And we recently were funded by, again, NIH Heart, Lungs, and Blood Institute R01 grant. So the idea is that we would use those enhanced 4D flow MRI measurements in patients to predict aneurysms that are at risk of growth and rupture. Expanding this method to other vascular territories, I started working on a biomedical device for capturing toxins from blood. So when liver tumors are treated by injecting chemotherapeutic drugs in the arteries feeding the tumor, a large fraction of these drugs pass into systemic circulation causing side effects. And the idea is to develop a device that can chemically bind and capture those drugs during the procedure. So from simple initial design, we went to a fairly sophisticated three-stage multi-channel design. And this work is collaboration with UC San Francisco, University of California, Berkeley, and Caltech. And it is funded by a National Cancer Institute R01 grant. So we conducted multi-scale and multi-physics computations coupling flow dynamics and transport and using analogy between heat and mass transport. And as you can see, we developed this honeycomb multi-channel configuration. We recently got a US patent for this device. My future directions are in studying models of flow cerebral spinal fluid in the brain. So that's analysis of brain health. And there are two directions. One is using deep learning approach to enhance imaging of cerebral spinal fluid that is removing metabolic waste from brain. So that's collaboration with Professor Ed Delp and Purdue Institute of Integrative Neuroscience. And another project is on finding biomechanical biomarkers predicting Alzheimer's disease, modeling brain ventricles. We have a pending R21 grant with Dr. Inge Tong. And I want to also briefly talk about teaching because that's the favorite part of my job. I teach biomedical transport to graduate and undergraduate students. And when I presented material to over 100 students with backgrounds in biology and engineering, I used dimensional analysis to show the similarity between the transport of fluids, mass, and heat. And I talk about different mechanisms of transport. You see a red blood cell navigating through the roads of circulation. And I talk about advection and diffusive transport. And so it's important that the students not just plug what's known to them into formulas that are set for them, but understand the physical meanings of equations, understand how to model physical phenomena and assumptions and limitations of the models and compare them to clinical data and measurements. And I was awarded attacker prize for teaching in BME for that. So I want to finish with acknowledging my external clinical collaborators from North Western University, Barrow Neurological Institute, UC San Francisco, IU School of Medicine, and University of Lyon and France. And of course, I wouldn't be able to do this research without my brilliant graduate students and postdocs. And I'm very grateful to them for their hard work. So thank you for your attention. And I'm ready for questions. Natalie, thank you. The question I always get when I see your presentations, and this is, I'm sure, the audience is thinking, how do you balance all these projects and how do you prioritize? Because someone with your talents, quite frankly, has lots of opportunities to make, in this case, biomedical contributions. Can you give us a sense how you weigh different research opportunities and maybe, and how that's changed over time in your career and in your learning process? Thank you, George. That's a fantastic question. That's probably the single most difficult problem, most difficult questions that I have in my research. And I think my collaborators and my students helped me very much. So I see what is interesting to my lab, to my team, where we can make progress. And my students explore multiple projects. And whenever we see an opportunity, something where we can really make a difference, we advance in this direction. So I'm really grateful to my grad students and postdocs for helping me prioritize. Yeah, that's what I found too, that diversity of opinion that also sometimes crosses generations can be tremendously valuable in terms of, and it brings a cohesion to the lab. Vitaly? Yes. I am going to ask Alice's question for her. What advice would you give your early self, given what you know now? Well, I would advise myself to remain confident and to get involved in data science as early as possible. Thank you. Data science is awesome. We understand how statistical methods can help us analyzing this data from models, measurements and experiments and how to fuse and blend this together and how to develop novel methods. Very good. Vitaly, thank you very much. Thank you. Thank you.