Eigenvector-based SPIRiT Parallel MR Imaging Reconstruction based on ℓp pseudo-norm Joint Total Variation

Abstract

Parallel Magnetic Resonance (MR) imaging is a well-established acceleration technique based on the spatial sensitivities of array receivers. Eigenvector-based SPIRiT (ESPIRiT) is a new parallel MR imaging reconstruction method that combines the advantages of the SENSE and GRAPPA methods. It estimates multiple sets of the sensitivity maps from the calibration matrix that is constructed from the auto-calibration data. To improve the quality of the reconstructed image, we introduced the Total Variation (TV) and ℓp pseudo-norm Joint TV (ℓpJTV) regularization terms to the ESPIRiT model for parallel MR imaging reconstruction, which were solved by using the Operator Splitting (OS) method. The resulting denoising problems with the TV and ℓpJTV regularization terms were solved by exploiting the Majorization Minimization method. Simulation experiments on two in vivo data sets demonstrated that the proposed OS algorithm with the TV regularization term (OSTV) and OS algorithm with the ℓpJTV regularization term (OSℓpJTV) outperformed the conventional method with the ℓ1 regularization term in terms of SNR and NRMSE. And the OSℓpJTV algorithm was slightly superior to the OSTV algorithm with the TV regularization term.