Rita G Nunes1,2,3, Joseph V Hajnal4,5. 1. Institute for Systems and Robotics and Department of Bioengineering, Instituto Superior Técnico, Universidade de Lisboa, Av. Rovisco Pais, 1, 1049-001, Lisbon, Portugal. ritagnunes@tecnico.ulisboa.pt. 2. Instituto de Biofísica e Engenharia Biomédica, Faculdade de Ciências, Universidade de Lisboa, Lisbon, Portugal. ritagnunes@tecnico.ulisboa.pt. 3. Division of Imaging Sciences and Biomedical Engineering, King's College London, London, UK. ritagnunes@tecnico.ulisboa.pt. 4. Division of Imaging Sciences and Biomedical Engineering, King's College London, London, UK. 5. Centre for the Developing Brain, King's College London, London, UK.
Abstract
OBJECTIVE: Point spread function (PSF) mapping enables estimating the displacement fields required for distortion correction of echo planar images. Recently, a highly accelerated approach was introduced for estimating displacements from the phase slope of under-sampled PSF mapping data. Sampling schemes with varying spacing were proposed requiring stepwise phase unwrapping. To avoid unwrapping errors, an alternative approach applying the concept of finite rate of innovation to PSF mapping (FRIP) is introduced, using a pattern search strategy to locate the PSF peak, and the two methods are compared. MATERIALS AND METHODS: Fully sampled PSF data was acquired in six subjects at 3.0 T, and distortion maps were estimated after retrospective under-sampling. The two methods were compared for both previously published and newly optimized sampling patterns. Prospectively under-sampled data were also acquired. Shift maps were estimated and deviations relative to the fully sampled reference map were calculated. RESULTS: The best performance was achieved when using FRIP with a previously proposed sampling scheme. The two methods were comparable for the remaining schemes. The displacement field errors tended to be lower as the number of samples or their spacing increased. CONCLUSION: A robust method for estimating the position of the PSF peak has been introduced.
OBJECTIVE: Point spread function (PSF) mapping enables estimating the displacement fields required for distortion correction of echo planar images. Recently, a highly accelerated approach was introduced for estimating displacements from the phase slope of under-sampled PSF mapping data. Sampling schemes with varying spacing were proposed requiring stepwise phase unwrapping. To avoid unwrapping errors, an alternative approach applying the concept of finite rate of innovation to PSF mapping (FRIP) is introduced, using a pattern search strategy to locate the PSF peak, and the two methods are compared. MATERIALS AND METHODS: Fully sampled PSF data was acquired in six subjects at 3.0 T, and distortion maps were estimated after retrospective under-sampling. The two methods were compared for both previously published and newly optimized sampling patterns. Prospectively under-sampled data were also acquired. Shift maps were estimated and deviations relative to the fully sampled reference map were calculated. RESULTS: The best performance was achieved when using FRIP with a previously proposed sampling scheme. The two methods were comparable for the remaining schemes. The displacement field errors tended to be lower as the number of samples or their spacing increased. CONCLUSION: A robust method for estimating the position of the PSF peak has been introduced.
Authors: Mark Jenkinson; Christian F Beckmann; Timothy E J Behrens; Mark W Woolrich; Stephen M Smith Journal: Neuroimage Date: 2011-09-16 Impact factor: 6.556
Authors: Dan Ma; Vikas Gulani; Nicole Seiberlich; Kecheng Liu; Jeffrey L Sunshine; Jeffrey L Duerk; Mark A Griswold Journal: Nature Date: 2013-03-14 Impact factor: 49.962