"Sparse Sampling and Constrained Reconstruction in Magnetic Resonance Imaging "
Speaker: Dr. Richard Frayne
Abstract: Magnetic resonance (MR) imaging is a powerful non-invasive tool in modern medicine and is used worldwide to support important diagnostic medical decisions. Building on earlier discoveries in nuclear magnetic resonance, the first MR scanner was commercialized in the early 1980s. These commercial instruments mainly adopted a data acquisition and image reconstruction strategy known as “spin-warp” or Cartesian imaging.[1] Spin-warp imaging was a critical development that uses a series of gradient magnetic fields to acquires MR data in k-space (the Fourier transform, FT, of image space). Image reconstruction is, in principle, quite simple, only requiring FT of the acquired k-space data to produce an MR image. Spin-warp imaging has a number of advantages including implicit averaging when using the FT and a predictable response to most image artifacts (e.g., motion, fat-water off resonance). Spin-warp imaging, however, has some drawbacks, primarily the lengthy time required to collect the entire k-space data set. Non-spin-warp imaging approaches have been under investigation since before commercialization of MR and have found specific application [2,3] in areas including fast, dynamic (or time resolved) and short echo–time imaging. These approaches include radial sampling (similar to computed tomography), spiral sampling and, most recently, sparse sampling. In this presentation, some sparse sampling approaches will be described and examples of their application presented. Their advantages and disadvantages relative to the spin-warp imaging stand will be discussed. Focus will be on the signal processing challenges in both data acquisition and image reconstruction when employing these methods.
[1] Edelstein WA, Hutchison JMS, Johnson G, Redpath T. Phys. Med. Biol. 1980; 25: 751.
[2] Mistretta CA. J Magn Reson Imaging. 2009; 29: 501.
[3] Kierans A, Parikh N, Chandarana H. Radiologic Clinics of North America. 2015; 53: 599.
Biography: Richard Frayne, PhD is a Professor (with tenure) in the Departments of Radiology and Clinical Neuroscience, a member and Deputy Director of the Hotchkiss Brain Institute (HBI, hbi.ucalgary.ca) and an associate member of the Libin Cardiovascular Institute of Alberta (LCIA, libin.ucalgary.ca), all in the Cumming School of Medicine at the University of Calgary. He directs the Vascular Imaging Laboratory (www.ucalgary.ca/vil) and from 2010-7 was the Scientific Director of the Seaman Family Centre, Foothills Medical Centre, Alberta Health Services (mrcentre.ca). From 2003-2013, he was a Canada Research Chair in Image Science and, in 2010, was appointed to the Hopewell Professorship in Brain Imaging. He is currently President-elect of the Society of Magnetic Resonance Angiography. He is also the Vice-Chair of the Heart and Stroke Foundation of Canada’s (HSFC) Scientific Review Committee.
Dr Frayne’s research interests are in the development and application of new imaging techniques and tools in humans for the study, detection and treatment of neurovascular disease. Current specific interests include angiography, perfusion/permeability and quantitative iron imaging in stroke and small vessel disease, and applications of these and other imaging techniques to human clinical trials. He is also interested in advanced image reconstruction and signal processing strategies.