Constrained Magnetic Resonance Imaging: Denoising and Sparse Sampling

Magnetic resonance (MR) imaging technologies have enabled new opportunities to reveal the mysteries of the brain and body -- their function and organization, and what goes wrong when they are injured or diseased. MR experiments are quite flexible, and the MR signal can be manipulated to noninvasively probe anatomy, physiology, and metabolism. However, while MR imaging is decades old and has already
revolutionized medical imaging, current methods are still far from utilizing the full potential of the MR signal. In particular, traditional MR methods are based on the Fourier transform, and suffer from fundamental trade-offs between signal-to-noise ratio, spatial resolution, and data acquisition speed. These issues are exacerbated in high-dimensional applications, due to the curse of dimensionality.

Classical approaches to addressing these trade-offs have relied on improved imaging hardware and more efficient pulse sequences. In contrast, our work addresses the limitations of MR using relatively less-explored signal processing approaches, which have recently become practical because of increasing computational capabilities. This talk concerns the use of constrained imaging models to guide the design of both data acquisition and image reconstruction, leading to improved imaging performance in the context of both noise-limited and resolution-limited scenarios. We will present and discuss some recent methods we've developed that incorporate constraints such as low-rank matrix structure, sparsity, and reference-based edge priors.

Justin Haldar received the B.S. and M.S. degrees in electrical engineering in 2004 and 2005, respectively, and the Ph.D. in electrical and computer engineering in 2011, all from the University of Illinois at Urbana-Champaign. He is currently a Research Assistant Professor in the Ming Hsieh Department of Electrical Engineering at the University of Southern California, where he is affiliated with the Signal and Image Processing Institute, the Dana & David Dornsife Cognitive Neuroscience Imaging Center, and the Brain and Creativity Institute. His research interests include image reconstruction, signal modeling, parameter estimation, and experiment design for biomedical imaging applications, with a particular focus on magnetic resonance imaging and spectroscopy. His work on constrained imaging has been recognized with a best student paper award at the 2010 IEEE International Symposium on Biomedical Imaging and the first-place award in the student paper competition at the 2010 international conference of the IEEE Engineering in Medicine and Biology Society.