Grad student Hernando wins Rabi Award for MRI research

ECE graduate student Diego Hernando won the Isidor I. Rabi Award from the International Society for Magnetic Resonance in Medicine for his paper and research on separating water and fat in medical imaging. Hernando is a student in ECE Professor Zhi-Pei Liang research group.

Hernando’s research was based on the problem surrounding medical imaging and the difficulty of separating the fat and water from magnetic resonance imaging (MRI). If the two are not separated, the imaging can look ambiguous, and it may be difficult to tell if what you are looking at is water or fat. Rather than trying to separate the images from one location, Hernando trying to separate these from the entire image using a special technique. “So we try to create the image for water and the image for fat for all the spatial locations simultaneously,” said Hernando. “It is a difficult problem mathematically, but a technique known as graph-cuts has enabled us to solve it very nicely.”

The graph cuts technique comes from computer science and is used to solve optimization problems. “It helps you find the best solution out of a very large set of possible solutions,” said Hernando. “For certain classes of problems, it is hard because your set of possible solutions is huge, and it would take you forever to look at each of them. So graph cuts are a tool for finding the best solution quickly.”

Hernando said that this area of research is significant and will help to more easily identify certain diseases, in particular, cardiac issues.

Hernando won the award after a two-stage competition where candidates submitted a full-length journal article, as well as a shorter conference abstract. Three finalists were selected based on their articles and abstracts, and they then presented their work in both a poster session and an oral presentation. Diego was selected as one of the three finalists early this year, based on his paper “Robust water/fat separation in the presence of large field inhomogeneities using a graph cut algorithm.” He won the award after the final round of competition involving poster and oral presentations of his work at the 17th annual meeting of the International Society for Magnetic Resonance in Medicine, held in Honolulu on April 20-24, 2009.

Other co-authors of the award-winning paper were Justin Haldar, Peter Kellman, and Zhi-Pei Liang. Haldar, also a Ph.D. student in Liang’s group, was an important contributor to the project. Kellman, a staff scientist at the National Institutes of Health (NIH) and research collaborator with Liang's group, brought unique experimental expertise and insight to the project.

Hernando is also excited about the future applications this imaging scheme could have. “I think that clinically, the biggest application is to reveal fat infiltration into muscles, like the myocardium,” he said.

Hernando is at NIH this summer, continuing his MRI research. “The basic techniques can be expanded in a number of different ways,” he said. He will begin looking at spectroscopic imaging, which can give a more precise characterization of fat and tissues. Such a capability can have a profound impact on biology and medicine.