By Fran Broderick
The following is part of a series of articles that highlight professors at the Golisano College. For our first installment we spoke with Dr. Vicki Hanson.
Dr. Pengcheng Shi is the Director of the Ph.D. Program in Computing and Information Sciences. He previously worked as a professor and researcher at Yale University, where his 1996 dissertation, ‘Image Analysis of 3D Cardiac Motion Using Physical and Geometrical Models’, and follow-up works helped define a new direction for computational heart image analysis. Dr. Shi joined the Golisano College in 2007 and helps lead dedicated students and researchers in continuing the advancement of cardiac imaging and computational cardiac physiology.
What has been the arc of tech during your career as a student and researcher?
Computing is still a relatively young field. If you look at computer science, fundamentally its roots are still in mathematics even to this day. But I think over the last 10-15 years, maybe a bit longer, we have started to see more other fields influence the direction of computing, or vice-versa. The way I see it, my first ten years – 1990-2000 – when I was a doctoral student and then a junior faculty it was more how I used computing to solve biomedical problems, or I should say turning biomedical problems into computing problems and then trying to solve them. But for the past ten years, starting around 2002-2003, it became more clear to me that kind of approach won’t work long-term. The reason is fundamentally I was solving a technical problem. I was not solving a biomedical problem. So my target audience – the physicians and the biomedical scientists who the work is meaningful for – don’t really give a damn. You can give someone equations and algorithms and data and images, but what do they mean? So, around 2002 I realized we are going to hit the wall pretty soon. We should find other ways that are closer to solving the real problems in order to help the targeted audience. We [computing researchers] can’t just feel good about coming up with some new fancy strategies and algorithms. So over the last dozen years or so, we are no longer computer scientists working on biomedical problems, we have become computational scientists. We now truly spend many hours, days, months or even years to understand the biomedicine – what are the problems, what they have done, what they have not done, what they want. And what the implications might be. You may be able to do it but more so, you need to be able to appreciate the why.
As far as research, it was love at first site - but I didn’t expect it to be. My mom was a famous chemist and my father was a university president. So the path was set for me to walk in many ways. But I really took to it once I was at Yale and began helping to bridge the gap between the cardiologists and the engineers and computer people.
Shi’s advisor at Yale held a joint appointment with the department of radiology. Therefore, computing researchers were working across the hall from cardiologists and other biomedical researchers. Although Shi had trained as an engineer, he had also studied biomedicine in his undergraduate years. This cross-disciplinary experience put him in a unique position and ultimately he worked with both teams, culminating in his graduate dissertation: ‘Image Analysis of 3D Cardiac Motion Using Physical and Geometrical Models.’ The dissertation represented a paradigm shift in computational medicine and announced Shi as one of the world’s foremost experts in cardiac imaging and computational physiology
What has your experience been like at RIT and where do you see the program going?
In 2008, we had 19 applicants to the Ph.D. program. This past year we had over 100. Last year we had a student - Zach Fitzsimmons - publish in a top-tier publication in his first year and win a NSF Graduate Fellowship (we have had three NSF fellows in the last four years). And a 2nd-year student just had her works accepted by two top conferences. Harvard, MIT, Stanford… most of these places, if you can get into one of these publications your Ph.D. is effectively done. But these are first year students publishing and presenting in major publications and at major conferences. Students have the ambition and they have the goals for themselves. That’s what’s really, really exciting. Any individual success is just noise. You need consistence.
Dr. Shi starts his Ph.D. Research Foundation course with an assignment asking students to cook their favorite meal. He then asks them to use someone else’s recipe to cook that same meal. Finally, he asks students to merge their favorite elements of the two meals, to form a further-improved third iteration. This approach is used to communicate the need for creativity and synthesizing of different information when researching.
What kind of advice would you give someone considering pursuing a Ph.D.?
Research is not different from many other things we do in life and in work, so people should not be afraid. Fundamentally though, you have to have the heart to do it. Research cannot be a job. You cannot say, “I want to be a researcher because it seems interesting.” That doesn’t necessarily work. Research doesn’t mean you have to be a genius, but you have to be passionate and creative. You have to be able to draw knowledge from many different fields. So, the difficulty then is how do you acquire information? How do you know who the giants are in related fields? I used to complain about my undergrad in biomedical engineering because I felt like a learned a lot without learning anything, because how deep and broadly you can go in the specific subjects is limited. About 10-12 years ago I realized those days in med school where I learned anatomy, biochemistry, physiology, etc. have helped me. I still need to relearn [these subjects] but at least I know they exist and what they are about.
[In research] I take your good part and your good part and somehow my creativity merges them. The way I see interdisciplinary research is you connect things. To not only see or borrow things but to connect them. To see why something might work in a given situation versus a lot of the time where we singularly look at our own thing. Don’t think you are smarter than others, because you’re not. But don’t think you’re not as smart as others and give up. If you have seen something others have not, it’s not because you’re smarter, it’s because you’ve read this and read that and put effort into seeing how it relates to what you do so you can integrate those methodologies and strategies. That’s what sets you apart. You’re not smarter - you’re better at integrating and standing on the shoulders of giants in different fields.