The development of advanced imaging, including digital x-rays, ultrasounds, MRIs, CT and PET scans, as well as new types of microscopy, has revolutionized biomedical and scientific imaging over the last half century, giving researchers and doctors powerful tools to explore the world and the human body.
But advanced imaging technology faces big challenges: Medical scans can be long, uncomfortable and prohibitively expensive for patients; scientific imaging applications are often too slow to fully capture chemical and biological processes; some types of imaging produce radiation that can interfere with their subjects; and many types of imaging are simply reaching the limits of how much data they can collect.
That’s why Ulugbek Kamilov, who starts at the University of Wisconsin-Madison as the Leon and Elizabeth Janssen Associate Professor in electrical and computer engineering in November 2025, uses computational imaging techniques and integrates physical models with machine learning to improve biomedical and scientific imaging applications.
Medical imaging is one area where Kamilov feels his work could have a big impact. For instance, he says patients undergoing chemotherapy or with other physical issues can have difficulty spending 20 or 30 minutes in a cramped MRI machine. That means the images produced are not as detailed as they could be. “Your treatment often depends on the quality of your scan,” says Kamilov. If his research could speed up the imaging process, it could make a real difference. “This is an area where my work can have a clear, positive impact on society. We can make some progress if we apply the technology in the right way.”
Kamilov earned his bachelor’s and PhD degrees from EPFL in Switzerland, focusing on signal processing optimization and machine learning for analyzing biomedical imaging data. There, he developed statistical models to improve the quality of accelerated MRI and CT scans.
He then spent three years at Mitsubishi Electric Research Labs working on imaging systems for autonomous vehicles before joining the faculty of Washington University in St. Louis, Missouri.
Over the last nine years, he has focused on machine learning and AI for biomedical and scientific imaging. Over time, engineers have developed different models of how imaging systems, like MRI, collect and interpret data based on the physics of the machines. “We know the physical models of how these machines work,” says Kamilov. “I integrate these models with machine learning models in a way that allows me to collect less data but to be more accurate and consistent.”
This allows the machines to do more with less, speeding up the imaging process. Kamilov often tailors his work to specific applications; for instance, he has worked with chemical engineers to maximize the speed of an MRI scan to capture chemical processes in real time. He has also developed MRI algorithms that accurately capture the motion of the heart and a CT scan that can image plants without giving them a lethal dose of radiation. Siemens has licensed some of his technology to accelerate MRI scans; he also worked for a year at Google applying his ideas to image restoration applications.
His research has earned Kamilov many awards, including an NSF CAREER Award and the 2024 IEEE Signal Processing Society Pierre-Simon Laplace Early Career Technical Achievement Award, which cited him for major contributions to theory and practice in computational imaging.
Kamilov says UW-Madison is an ideal place to continue his work due to the strong machine learning optimization and AI research communities. He’s also excited to collaborate with colleagues in radiology and the medical school as well as researchers in microscopy, bioimaging and scientific imaging. “It’s a very diverse scientific community that is hungry for advanced imaging technology,” he says. “And I’m excited to contribute and collaborate with them.”