Artifact correction in fast magnetic resonance imaging

Abstract

Magnetic resonance imaging (MRI) has been widely used for clinical diagnosis and preclinical studies. The use of phase-array coils has brought unprecedented opportunities for artifact correction and parallel imaging reconstruction by exploiting its capability of providing additional spatial information. The main scope of this thesis is to provide robust artifact correction for fast MRI in presence of acquisition imperfection or undersampling. First, a robust reconstruction method with k-space implementation and Nyquist ghost correction was proposed for simultaneous multislice (SMS) echo-planar imaging (EPI). 2D phase error correction SENSE (PEC-SENSE) was recently developed for Nyquist ghost correction in SMS EPI reconstruction where virtual coil simultaneous auto-calibration and k-space estimation (VC-SAKE) was used to remove slice-dependent Nyquist ghosts and inter-shot 2D phase variations in multi-shot EPI reference scan. However, masking coil sensitivity maps to exclude background region in PEC-SENSE and manually selecting slice-wise target ranks in VC-SAKE are cumbersome procedures in practice. To avoid masking, the concept of PEC-SENSE is extended to k-space implementation, and termed PEC-GRAPPA. Furthermore, a singular value shrinkage scheme is incorporated in VC-SAKE to circumvent the empirical slice-wise target rank selection. The proposed PEC-GRAPPA approach effectively removes the artifacts caused by Nyquist ghosts in SMS EPI without cumbersome tuning, providing a robust and practical implementation of SMS EPI reconstruction in k-space with slice-dependent 2D Nyquist ghost correction. Second, a novel method for EPI Nyquist ghost removal, which employs a k-space-based phase error estimation and performs the EPI Nyquist ghost correction by incorporating iterative self-consistent parallel imaging reconstruction (SPIRiT), termed PEC-SPIRiT, was presented. Aforementioned PEC-SENSE relies on pixel-wise operation in image-space, which is sensitive to distortion mismatch between calibration scan and EPI scans and localized pitfalls in phase error estimation. In the proposed method, both phase error estimation and correction are performed in k-space, and the phase error estimation can be updated iteratively, leading to gradually improved reconstruction. The proposed method was evaluated using both phantom and in vivo studies, demonstrating its robustness against distortion mismatch and high acceleration. Third, a novel calibrationless parallel imaging reconstruction algorithm was presented. Autocalibrating parallel imaging requires sufficient autocalibration signals (ACS) for reliable estimation of coil sensitivity. However, this is not feasible in some applications, for example, spectroscopic imaging where matrix size is relatively small. Recent publications proposed to construct k-space data into block-wise Hankel matrix, and perform parallel imaging reconstruction via low rank matrix completion. The proposed approach constructs a block-wise Hankel tensor instead, and uses tensor completion techniques to synthesize the unacquired samples. This method can also be extended to reconstruct multiple slices simultaneously and provide more accurate reconstruction. In summary, these studies have demonstrated that parallel imaging can provide a robust solution for artifact correction arising from imperfect acquisition or undersampling in fast MRI.