Making A Real Impact - Profile Casey Stowers

Casey Stowers is a third year CSEM student at the Oden Institute. She was recently awarded a National Science Foundation Graduate Research Fellowship Program (NSF-GRFP). This provides Stowers with three years of funding ($34,000 stipend and $12,000 for the cost of education each year). She is using it to work in Tom Yankeelov’s Center for Computational Oncology where she is designing models that would predict the response of breast cancer patients to neoadjuvant therapy.

Stowers grew up primarily in central Illinois, but spent a good portion of her life moving around due to her father’s job working for Caterpillar. “I’m mostly from Illinois, but my family moved a lot when I was a kid, so I've lived in Illinois, Colorado, Dubai and Belgium,” she said.

In high school, she lived between Argonne National Lab and Fermilab, the national particle physics and accelerator laboratory, which allowed her to participate in research at Argonne and attend several talks at Fermilab.

Stowers then attended Purdue University where she completed her BSc in Industrial Engineering.

“I studied industrial engineering at Purdue because that was the most computational engineering they had. Generally, the professors who were doing optimization or computation type research were all in that department.”

During her undergraduate degree Casey worked with a professor in mechanical and biomedical engineering on reconstructive skin surgery modeling at Purdue. “That's where I got really into research,” she noted.

She was also a computing intern for two summers at Lawrence Livermore National Laboratories (LLNL) working on deep learning models for optics inspection at the National Ignition Facility.

Computational science and engineering appealed to Stowers because of its applications in a variety of different fields, ranging from environmental science to medicine. “I like that I can make an impact in almost any field without having to be an expert in that particular subject.”

“I applied to the CSEM program at the Oden Institute because I enjoyed doing computational and medical work during undergrad, and this was one of the only PhD programs that would let me keep doing both computational and medical.”

“Although I’m interested in many different things, I think medical currently interests me the most because I feel like it has a really tangible impact,” said Stowers, adding, “When I'm working with data, I work with patient data and my research is working directly with doctors. I’m not just publishing a paper and never touching the work again—my work will have a real impact on cancer patients.”

Stowers works with her advisor, Tom Yankeelov, in the Center for Computational Oncology where she is exploring methods for incorporating genetic data into a multi-scale mathematical model of tumor growth designed to predict the response of breast cancer patients to neoadjuvant therapy.

“The ability to predict the response of cancer patients could allow for intervention and therapy optimization for the individual patient, which would limit their exposure to toxicities associated with unsuccessful therapies,” she said.

“When a patient comes in for therapy, this is a triple negative breast cancer patient, they do something called neoadjuvant therapy. They have imaging before they start the treatment, then they do two cycles of the treatment, then another set of imaging, two more cycles of treatment, another set of imaging, the second type of treatment, another set of imaging and then surgery,” explained Stowers.

“The previous work has been able to take the first two cycles of imaging and calibrate a PDE model in order to predict the third cycle of imaging, but if we have to wait until they're halfway through the therapy before we know if they're going to respond, then they've still done half the therapy.” 

“My research asks, what if we take what we have at baseline - at the very first imaging time point before therapy has begun - and we try to predict what happens at the end? This means I only have a couple hundred data points. Convolutional neural networks are generally on the order of tens of thousands of data points for training, so I don't have a lot to work with.” 

Stowers is proud of many of her achievements throughout her academic career, including her internships at LLNL, her publications—including one in a clinical journal—and, of course, her NSF fellowship.

“It was really cool because my sister also won the same fellowship, so we have two people in the family to get an NSF fellowship.”

“For my application, I focused on the broader computational impact of my research,” Stowers explained. “While I’m working with medical data, this project is impactful for any areas where you have limited data and/or data that can be difficult to connect to underlying physical processes, making it difficult to model.”

When not working on her research, Stowers enjoys musical theater, hiking and playing D&D.

She also spends a lot of her time knitting or sewing her own clothes. “I met with another CSEM student and I got her into knitting and now we knit together. It's a very addictive hobby. You can also watch as much TV as you want and not feel bad about it.”

Stowers and her husband have two guinea pigs. “Their names are Minnie and Ollie. We were going to get two more, but they are two old grouchy ladies and they did not want friends,” she said.

As a third-year student, Stowers is excited to be finished with classes so that she can focus all of her attention on her research. “I made a lot of progress this summer and I look forward to continuing that progress this fall.”