Release Time:2019-03-12  Hits:

Indexed by: Journal Article

Date of Publication: 2014-01-01

Journal: Journal of Computational Information Systems

Included Journals: Scopus、EI

Volume: 10

Issue: 5

Page Number: 2009-2016

ISSN: 15539105

Abstract: l1-SPIRiT is an efficient reconstruction method for partially parallel magnetic resonance imaging (pMRI), which introduces wavelet-based sparsity into the SPIRiT framework. To exploit the fact that coil images are sensitivity weighted images of the original image, a joint-sparsity model is proposed in the l1-SPIRiT method. Coil images are transformed into sparse coefficients and then coefficients from different coils at the same spatial position are jointly penalized to exploit inter-coil similarities. In this work, the sparse representation is based on singular value decomposition (SVD), which preserves different levels of characteristics for the coil images in an efficient way, additionally, its adoption into the joint-sparsity model of the l1-SPIRiT method opens a new way to make use of the inter-coil structure similarities. The performance of the proposed method was tested on two datasets with various experimental settings. The experimental results indicate that, compared to the original l1-SPIRiT method, the proposed method is more robust and efficient. ? 2014 Binary Information Press.