Students Profiles

Meet this year’s MathWorks Fellows: A three-part series, II

The 2020-21 School of Engineering MathWorks Fellows use MATLAB software and engineering know-how to tackle problems in modern medicine

Kate S. Petersen

The cardiologist reviews the diagnostic images as they play across her screen. Her patient has developed a dangerous lesion in his vasculature and she must implant a stent in the area to return proper blood flow. She stops the stream to get a better look, and then lets it play forward again. There are thousands of images stitched together that, collectively, paint a picture of the lesion otherwise hidden from her view. What is it composed of? How big is it? How deep into the tissue does it intrude?

Unfortunately, the image stream is imperfect, explains Max Olender, now a postdoctoral fellow in the MIT Edelmen Lab who was awarded his PhD from MIT earlier this year. Both the images themselves and the manual analysis process may fail to reveal crucial details about a given patient’s condition, details that directly inform the success of any medical intervention. Olender is putting machine learning to work on the problem.

“[During my PhD, I worked] towards developing better…models of diseased coronary arteries using clinical imaging,” he says. “We get these images from clinical partners that show the diseased vessel and I've developed several tools…to extract key information…that allows us to better understand the state of the disease for that given patient.”

Olender’s PhD research was supported by a MathWorks fellowship. MathWorks is a software company that was founded in 1984 with the goal of providing researchers with new and more powerful computational tools. In 1985, the company sold ten copies of its first product, MATLAB, to MIT. MATLAB is a programming language that can be used to develop algorithms, analyze data, and build mathematical models. Almost forty years later, MIT students and faculty continue to wield MATLAB, and another MathWorks product called Simulink, to advance research across disciplines.

In 2019, in celebration of the decades—long partnership with MIT, MathWorks began awarding fellowships to graduate students in the School of Engineering who utilize MATLAB and Simulink in their research.

Olender uses MATLAB machine learning, deep learning, and physics-based image processing features to detect and quantify variations in vascular lesions between patients.

“Then we can look at the relationship between those quantitative features that we've derived and things like clinical outcomes,” says Olender. This process could eventually help cardiologists more effectively tailor their interventions to a specific patient’s unique presentation of vascular disease.

In addition to the clinical applications of scientific research, Olender is interested in the ways that science informs social policy and vis versa. As such, he has led the MIT Science Policy Initiative, serves on the Journal of Science Policy & Governance Editorial Board, and will begin a Biophysical Society Congressional Science and Technology Policy Fellowship this year. He also tries to get out hiking in the White Mountains of New Hampshire whenever he can.

Sırma Örgüç, a MathWorks fellowship recipient and postdoctoral fellow at the MIT Institute for Medical Engineering & Science, is also interested in dedicating her engineering and materials science expertise to medical applications.

“I am essentially trying to be at the intersection of engineering and medicine,” says Örgüç, who, like Olender, accepted her current position after completing her PhD at MIT. “[I am] trying to communicate with the medical experts as much as possible to…understand what kind of questions they have [and how] engineers can help.”

While she enjoys cooking, getting outside for long bike rides with friends, and even holds a teaching certificate in piano, she spends long hours in the lab designing hardware and software systems that can support medical efforts targeting different parts of the body.

For instance, she has also developed artificial muscle fibers with applications for both human prosthetics and soft robotics. Örgüç says that, while this was a particularly difficult effort, the resulting research paper made the cover of Science.

During her PhD, Örgüç’s MathWorks fellowship supported a neuromodulation project in which she designed probes that were implanted into the brains of mice. The probes both recorded brain activity and stimulated it using LED lights. The system was controlled by a MATLAB-based interface, which also logged experimental data. Now as a postdoc supported by a Schmidt Science Fellowship, she is working on testing and developing related software and hardware systems for neuroscience applications that, in the future, could be used to help control levels of consciousness in surgery patients or to treat neurological problems such as Parkinson’s disease or epilepsy.

For Örgüç, the potential applications of engineering and materials science in the biomedical field is incredibly broad and could include novel implantable or ingestible systems, futuristic operating room hardware, or improvements to the size and usability of existing medical tech. But whatever the application, the use of MATLAB seems to be an enduring theme.

“I have been using MATLAB since [undergrad],” says Örgüç. “So, it's been part of all my research projects and it looks like it's going to be for a while.”



content Link link