 This is a really great event for a lot of us because we have a chance to celebrate the success of our rising superstars, a recently promoted associate professor, so it's really a point of pride and joy for us because really the future of Purdue Engineering is in the hands of these outstanding individuals that are currently in this position. The Celebrating Associate Professors in Engineering pretty much started, was it last year, and the idea is not to go too deep into a technical area but it's really an opportunity for these colleagues to share with the rest of us an insight into like some decisions they may have taken that led to their success, what are the secrets of success for them, and a chance also to of course then to reach out to you know what to other colleagues certainly to look for new collaborations and the third important piece is to really for us to get a sense of what the vision is going to be, what great things they're planning for in the coming years. So thank you all for coming and without further ado I'm going to invite David Humilis the Associate Head of the School of Biomedical Engineering to introduce our first speaker for the day. Thank you very much for this opportunity to introduce one of BME's distinguished faculty. I've gotten to know Craig over the years, I remember having a welcome to Purdue introduction to a martini at red seven shortly after he arrived and ever since then I've gotten to see his program take off and it really has. Coming from Minnesota then to wash to wash you for a PhD before going to Stanford for PhD and then postdoc at Harvard in the space of technology and cardiovascular health. I hope today he'll talk about four-dimensional ultrasound and some studies in preeclampsia among other things moving forward. He's won a number of awards and I had to make a little cheat sheet to keep track of just a few of the big ones including an award the Rita Schaefer BMES young investigator award and a faculty service award outstanding mentorship award from BMEGSA and also a number of financial awards to support his research including a large grant from the American Heart Association and the HR 21s and multiple pages of awards that fund his research program. So without much more introduction he has 20 slides to give in 10 minutes. Let's get going thank you. Thank you David thank you for coming everybody it's great to see everybody can everybody hear me okay good all right so I will plop over to our laptop fantastic right so um I'll be talking a little bit today about the work that's happening in the cardiovascular imaging research laboratory I see a lot of friendly faces in the back these are all the students that are making this work possible and I wouldn't only be able to talk about half of this if they weren't here doing all this great amazing research um I'll talk a little bit about the the work we do regarding strain in relation to both cardiac and vascular disease some compositional information in imaging we do not only with ultrasound but photocoustic technology and then some device development whether it's an app for preeclampsia or in collaboration with Chi Wan Li who's another professor of biomedical engineering at cardiac patch so we've got multiple areas all kind of centered around cardiovascular disease which is our focus although we are collaborating with other groups in oncology and mechanical engineering and other areas trying to think about different applications but we use small animal imaging to understand disease progression using an engineering approach so one of our first projects I'll highlight is work with David Reuter and George Wattica David is a pediatrician out in Seattle and he came to us talking about the problems of preeclampsia so preeclampsia is hypertension induced during pregnancy it affects about three to seven percent of all pregnancies in the united states there's about seventy thousand maternal deaths a year worldwide and five hundred thousand fetal deaths worldwide every year due to preeclampsia um it starts around the 20th week of pregnancy and in reality a lot of women don't know they have this until pretty severe symptoms occur or in developing countries they're basically going into labor at early stages preterm birth is a big problem with preeclampsia because of this so we're working with clinicians both in seattle as well as our our contacts in obi gyne clinic down at iu hospital um university hospital in indianapolis to develop an automated supine pressure test the idea here is that you can take blood pressure in different body positions and with the woman lying on her side and then on her back if there's a big change in blood pressure a big increase that woman is at a much higher risk for developing preeclampsia later on during pregnancy so we're trying to develop an app with a cuff a smartphone that can record her body position record her blood pressure use that data to then send off remotely to her clinician to not have to have her come into the hospital or to a remote clinic every week every two weeks and identify who's at risk for developing preeclampsia earlier on so the treatment can happen before side effects become uh too late so our first publication just came out hamna koreshi is the master student that led this work we're just submitting a phase two gates award the initial work was funded with a phase one grand challenge award from the gates foundation and we're targeting not only rural populations in indiana but sub-saharan africa india and other developing countries another project is with dave uh capillari capillari i think yeah uh and louis solario uh it actually started from one of these associate professor talks that dave gave uh a year or so ago and we were we were talking i knew dave just from being around the campus but only learned about his research once he he gave a talk about his little micro robots micro machines however we want to uh describe it and the neat thing about these are that if you have a magnetic field rotating these tiny little robots these are you know about a millimeter in length can move depending on the changes in the magnetic field and so he had this great technology and we came and said what we've got access to animals we've got this nice ultrasound approach we can like actually visualize things so this is the mouse colon we can do some preparatory work and then in the animal apply the same magnetic field and watch how this moves up and down so this is still pilot work we're in exploratory phases but there's a whole bunch of now interesting applications for oncology with drug delivery with biopsies with thinking about clearing the colon with um targeting different aspects so we're in talks with different uh mds in oncology through iu and other applications thinking about where this technology can head but we're excited it was initially presented at a conference in finland just this past summer working on a manuscript and grant publications from this work as well i'll talk a little bit more about some of the imaging work but um j humphrey is a professor at Yale that helped us with some of our um simulation work with vascular ultrasound so we can do really high resolution mouse aneurysm imaging with ultrasound and then combine that with some x vivo optical techniques build these three-dimensional models look at flow and and the true lumen and this complex false lumen and then look and actually create simulations so this is a cfd a computational simulation of blood flow through an aneurysm at different uh levels of severity some very simple uh uh expansions all the way to these multi-lobed dissections and what you can see here is this recirculating flow and the fact that we can combine our imaging approach with some computational techniques allows us to look at while shear stress and other metrics that you can't get just from direct measurements themselves so this is allowing us to think about progression severity when aneurysm patients need to be treated and why aneurysms continue to grow to rupture so with all that i wanted to now focus primarily on um coronary artery disease because i know david wanted to hear more about four-dimensional ultrasound so that's what i'll be talking about the rest of the talk heart disease is a big problem over a million patients the united states have a heart attack every year you get build up in your coronary arteries leading to blockage of flow and then ischemic disease you actually have a heart attack right so bernie sanders this just happened to him and we're trying to figure out now is he in this group of patients that have adequately healed infarctions or a low number of microfiber blasts inadequate healing and then severe dilation and then is he part of this 20 percent of patients that have a heart attack to develop heart failure within five years right now we don't know that you can do ultrasound you can assess cardiac function ejection fraction stroke volume with these different rules but each of these are 2d approaches so whether you're doing it with a 1d tracking over time or a 2d view of the whole left ventricle you have to assume geometries with the simpsons approach or a tight holds approach that aren't really accurate so what the left ventricle really looks like so what we've done is taken our animals our mice it's tiny little 25 30 gram mice anesthetize them monitor ecg and respiration we then can place an ultrasound probe on the chest wall take a bunch of short axis images and then rebuild this the the actual data based on where the transducer is in space so with gating for both ecg and respiration we can build four-dimensional ultrasound so a better way to actually see this is a schematic here with a 40 reconstruction we can take one location and a little bit further up and a little further up until we get all this data and then reconstruct one volume that represents right at the peak of the qrs complex a little later a little later a little later so we can reconstruct all this into four-dimensional data right gated volumetric information showing the volume changes over time what we can get is something looks like this long axis short axis you get a lot of these views with ultrasound but this coronal view where you see the sternum and rib cage artifacts is actually something very unique to this approach that other techniques don't have you can see that we don't have a sphere or an ellipsoid but something that looks much more complicated when you actually look at the geometry we then did a model of permanent legation so arvin supratna is in the back here he did a model where he can go in and actually open up the chest wall and put a tracheal tube in the mouse tie off a coronary artery and cause a heart attack in these small little animals we monitor their pain abby don't worry we can't we take good care of her mice with everything is approved by the Purdue animal care and use committee and we're very considerate of what's happening to these mice but after this right there's thinning of the wall this this big dynamic muscle turns into scar tissue it's not moving nearly as much and you can see it down here even at peak systole there's a much larger end systolic volume with that approach we can think about then strain so using some collaborations and thinking about from a mechanical engineering approach how does the wall deform you can see in the remote regions here we see high strains but in this lower region we get much lower strains suggesting that the wall is not nearly as dynamic it's not moving nearly as much so what arvin did was he did a study with 15 animals permanent legation where he tie off the artery completely a scheming reperfusion where we tie off and then have the flow come back and then a sham procedure where we basically just put a suture underneath and don't tie off any of the flow through the coronary we'd imaged he imaged at 1 2 3 5 7 14 21 and 28 days so this is a lot of work on his part he just got the paper accepted last week right before his defense so it worked out perfectly for timing and we were doing four-dimensional ultrasound some Doppler ultrasound for flow and then impulse wave Doppler assessment as well for diastolic dysfunction tissue harvest sacrifice and then looking with a histology at the end this is kind of what we get sham procedure even all the way out to day 28 you see healthy functioning normal myocardium ischemia reperfusion we get these sub sub endocardial apical infarcts where it's just a small portion of of the apical region that's affected and then in the permanent legation group you get this transmural infarct with a large amount of dilation that's evolving over these four week period so that's an example of what the ultrasound looks like in just 2d slices but now we can make these beautiful strain maps again healthy we get about 40 45 strain with the ischemia reperfusion it's much lower but only in the apex or in this for part down here with a little bit of remodeling and in the permanent legation group you get much reduced and much larger apical dilation and thinning with a much lower strain values and if we lay these out if you actually take a look at the apex laid out on a bullseye map you can see the four different groups here baseline sham ischemia reperfusion and permanent legation the white line here represents where the wall is thinned where the black is the strain estimated assessment of where we think the infarct is that kind of that border zone between healthy and diseased areas i'm getting a cue here and i'm wrapping up real quick so the wall thinned estimate basically this is what everybody else in the field is doing where's their thinning where's the infarct and you can see how this changes over the foray period but the neat thing that arvin and i was able to identify was that even at day one with this strain estimated approach we can see that we get about 40 percent of the left ventricle is infarcted or damaged based on this heart attack so we think with this approach we can use this strain estimated boundary to predict the eventual extent of wall thinning or how big the infarct is going to be four you know four weeks later so this is accelerated mice obviously in humans we're talking about months years later and identifying which patients we need to go more aggressive with therapy and other ones that are going to be fine and don't need as much therapy from there so some takeaway message uh david's happy to talk about 40 ultrasound as well the 3d estimation of strain allows us to in vivo quantify 3d strain with these animal specific geometries and strain maps we can monitor disease progression and the role of strain capture unique remodeling uh based on the severity of the infarct the sigmoidal pattern between the the infarct itself in the remote zone allows us to estimate the infarct boundary size and eventually right we want to be able to identify this new strain to identify patients that are at risk for heart failure after heart attack and think about improving patient care post prognosis after this event has happened so with that again i need to thank the group this is a wonderful group of students staff postdocs that have been a huge help in the work that i presented here our collaborators both at purdue externally one of the things i love about purdue is that i've come thinking about just a few things and then as you talk to other areas and other people you you give all these other collaborations and areas you never knew about i also want to thank my purdue mentors and colleagues starting off here just starting a lab you never know if what you do is the right approach but these are some senior people that have been a huge help for me getting the lab off the ground as well as two other additions recently teaching me humility and patience so with that thank you for your attention we have five minutes for questions because it's recording we'd like to hand this to you hi very nice work thanks so i'm really interested in sort of the preeclampsia work that you just got at the beginning you've done some looking at risks for pregnancies at 20 months and beyond have you looked at before pregnancies risks and looking developing apps or something to yeah determine that so actually the first study that we published was with healthy graduate students right now in purdue that we're not pregnant and assessing just what does body position do to blood pressure measurements before pregnancy or anything else right and what we found is that if you have a woman lie on her side and then move to her back just normally it's a 10 millimeter mercury increase and right now in the clinic it's really not clear what body position does or has an effect on blood pressure when you would think about 120 over 80 is your blood pressure but nobody ever says well that's actually lying on your back or lying on your side or sitting up or standing up and there's a big difference in body position even 10 millimeters of mercury that people are making big clinical decisions whether to be put on anti-hypertensive medications and a whole bunch of other things so that's our first control group our next group this study that did follow on was that 20 weeks gestation and later women that didn't have preeclampsia and we were seeing about a 14 to 15 millimeter mercury increase in those patients so below this 20 millimeter threshold that we're targeting and our next study is with patients that are going to be inpatient in Indianapolis at the University Hospital I think they have about 350 patients a year they get brought in for hypertension during pregnancy that we can now see those are the ones that either have preeclampsia are about to develop preeclampsia and if we get this increase of 20 or more uh millimeters and mercury based on on changes in body position but you can think about not only pregnancy but obesity and other changes in abdominal mass and how that affects not only the hypertension in general but renal blood flow which is kind of the underlying mechanism we're targeting at this point but great question I was wondering something so the results are very impressive so what would it take for this type of study and work to actually become something that when we go to the clinic it will be something common and uh used yeah what kind of hurdles do you have to go through before it becomes accepted for the the 40 ultrasound specifically the the 40 ultrasound work specifically yeah so the question is about how does this mouse 40 ultrasound move into the clinic or how does that have an impact in patient care um so yeah I think there's some inherent differences right the transducer on this mouse is between basically the the left side all the way across the sternum so when you're doing ultrasound in a human you're talking about much smaller probes that can go through different windows right between the ribs underneath the rib cage right these cardiac notches or windows so they're now making much better 2d arrays and there's a Jen Anderson's a student in the group a sonographer that's used a few of these where you can get not only one slice but whole volumes and now the question is can we get volumes that are fast enough with a high enough resolution that we can acquire enough of a volume of a region of interest maybe we need two or three or four of these that we then stitch together later on to put one big volume together but that's kind of what we're doing with the slice by slice approach with our mouse work so I think a lot of the analysis after we have the data acquired could be applied directly to what's happening now with these 2d arrays in the clinic great question so how long does the combination of 4d how long does the combination of 4d ultrasound imaging and also 3d strain images take for each little animal so the question then is how long is acquisition and analysis take arvin well how long do you spend doing this it's a no big deal right I think if we're going to go into the clinic and actually make this a workable strategy it's going to have to be faster so we're talking about machine learning and AI approaches to make the segmentations quicker increase the strain code efficiency where instead of having to wait three four hours it's much more automated and it's going to run in the background so that it might not be five minutes but we can't wait three hours between acquisition and diagnosis right and then I do have to do like multiple imagines during a certain period of time like a month or so after the heart attack yeah so the question is when do you do the imaging for the patient so you can and I think that's the big thing with these patients that have heart attacks is initially they can do an assessment of ejection fraction but they know it's going to change over time and so in the clinic yeah they're asking patients to come back in after a month after six months after 12 months to look what's what the progression is and if they are going down this route for heart failure can we improve their medication can we think about ventricular assist devices or are we thinking about eventually transplants thank you great all right I have one more question okay so as Leslie Gettys associate professor and he was always a such a forward looker a forward thinker in biomedical engineering biomedical technology now that you can take more risks as an associate professor post tenure where do you see this heading next and how is that inspirational for how you think about this moving forward yeah so thinking about impact I think is where we're headed and I think we're thinking about translation in the clinic but also there's there's impact can be broadly defined and we're talking to biotech and pharma companies obviously the lily connection here thinking about what we can understand from a disease progression animal model standpoint is helpful I also I have this background in cardiac and vascular disease that we're passionate about but the the women's health and reproductive health is something that we're now exploring more with this pre-eclampsia project another paper just came out in collaboration with a group at Tulane where we're looking at cervical insufficiency and changes in the cervix during pregnancy and during labor and delivery there's a lot of biomechanics work that's been applied to to hearts and vessels that could be applied to reproductive system that hasn't been done before that is really important for pelvic organ prolapse preterm birth a whole bunch of other things so we're exploring new areas and excited about different projects all right thank you again dr. gorgin thank you