A flexible two-dimensional phase correction for interleaved echo-planar imaging reconstruction

Abstract

Interleaved echo-planar imaging (EPI) is a fast clinical magnetic resonance imaging (MRI) scheme that obtains multiple echoes with a proper phase encoding (PE) strategy to generate multiple k-space PE lines. Since these PE data lines are from different echoes that may carry different phase and amplitude errors originating from the static magnetic field inhomogeneity and nuclear spin relaxation, to form an image free of artifacts, both phase and amplitude errors need to be compensated properly. To address this issue, we have developed a general image reconstruction technique which is capable of accomplishing two-dimensional (2D) phase correction for image reconstruction of interleaved EPI data. In this technique we formulated the reconstruction as a problem of finding an optimal solution to a set of linear algebraic equations corresponding to an imaging measurement. The phase errors, as well as other constraints, can be incorporated into these equations. The final solution can be obtained by inverting the coefficient matrix of the equation via a complex singular value decomposition (SVD) procedure, free of k-space data gridding. 2D phase corrected images have been successfully reconstructed using a set of imaging data acquired on a clinical MRI scanner. The significance of the work is that it has demonstrated that the 2D spatial phase correction can be accomplished for a set of interleaved EPI acquisition. Also, this is a flexible image reconstruction method for further improving the resulting image quality.