Efficient Deblurring by Iterative Reconstruction for Spiral-in-out Imaging

H. Guo, Q. Zhu, T. B. Harshbarger, A. W. Song Department of Biomedical Engineering, Duke University, Durham, NC, United States, Brain Imaging and Analysis Center at Duke University, Durham, NC, United States, Electrical and Computer Engineering, Duke University, Durham, NC, United States Introduction To address the signal loss due to the susceptibility differences at the air/tissue interfaces, several methods have been proposed to minimize the intra-voxel phase dispersion and recover the lost signal. Among them, the combination of spiral-in and spiral-out sequences [1, 2] has been proven effective and efficient in improving image quality. However, due to the long read out window of these single shot acquisitions in the presence of large field inhomogeneity, the offresonance effect can be severe which would induce noticeable blurring artifact in the final images. The conventional method to de-blur the images includes measuring a field map using extra scans and reconstructing images based on the region-specific frequencies [3], which tends to be time consuming even if a coarse field map is used. In this abstract we propose to use the intrinsic field map from the spiral in and out acquisitions and an iterative reconstruction method [4] to correct the off-resonance effect.