Victor B Xie1,2, Mengye Lyu1,2, Ed X Wu1,2. 1. Laboratory of Biomedical Imaging and Signal Processing, The University of Hong Kong, Pokfulam, Hong Kong SAR, China. 2. Department of Electrical and Electronic Engineering, The University of Hong Kong, Pokfulam, Hong Kong SAR, China.
Abstract
PURPOSE: To develop a new Nyquist ghost and geometric distortion correction method in echo planar imaging (EPI) using parallel imaging. METHODS: Two frames of EPI data are acquired with normal and phase-labeled sequence. The phase label is applied by modifying the PE prephase gradient to shift the central echo by one echo spacing. GRAPPA weights are trained from both frames and used to reconstruct images from positive or negative echoes in each frame to remove Nyquist ghost. Geometric distortion is then corrected by the B0 field map generated from the phase difference between positive and negative echo images. Phantom and in vivo experiments at 7 Tesla (T) and 3T were performed to evaluate the proposed method. RESULTS: Nyquist ghost was greatly reduced in all images even under oblique imaging and poor eddy current conditions, yielding significant improvements over the existing reference scan and image entropy minimization based methods. Image geometries were fully restored after distortion correction. Phantom results indicated that the signal-to-noise ratio efficiency was largely preserved while fMRI results showed no apparent degradation of temporal resolution. CONCLUSION: The proposed method provides robust correction of both Nyquist ghost and geometric distortion in EPI, and it is particularly suitable for dynamic EPI applications. Magn Reson Med 77:1749-1761, 2017.
PURPOSE: To develop a new Nyquist ghost and geometric distortion correction method in echo planar imaging (EPI) using parallel imaging. METHODS: Two frames of EPI data are acquired with normal and phase-labeled sequence. The phase label is applied by modifying the PE prephase gradient to shift the central echo by one echo spacing. GRAPPA weights are trained from both frames and used to reconstruct images from positive or negative echoes in each frame to remove Nyquist ghost. Geometric distortion is then corrected by the B0 field map generated from the phase difference between positive and negative echo images. Phantom and in vivo experiments at 7 Tesla (T) and 3T were performed to evaluate the proposed method. RESULTS: Nyquist ghost was greatly reduced in all images even under oblique imaging and poor eddy current conditions, yielding significant improvements over the existing reference scan and image entropy minimization based methods. Image geometries were fully restored after distortion correction. Phantom results indicated that the signal-to-noise ratio efficiency was largely preserved while fMRI results showed no apparent degradation of temporal resolution. CONCLUSION: The proposed method provides robust correction of both Nyquist ghost and geometric distortion in EPI, and it is particularly suitable for dynamic EPI applications. Magn Reson Med 77:1749-1761, 2017.
Authors: Vince D Calhoun; Tor D Wager; Anjali Krishnan; Keri S Rosch; Karen E Seymour; Mary Beth Nebel; Stewart H Mostofsky; Prashanth Nyalakanai; Kent Kiehl Journal: Hum Brain Mapp Date: 2017-07-26 Impact factor: 5.038
Authors: Elizabeth Powell; Torben Schneider; Marco Battiston; Francesco Grussu; Ahmed Toosy; Jonathan D Clayden; Claudia A M Gandini Wheeler-Kingshott Journal: Magn Reson Med Date: 2022-07-25 Impact factor: 3.737