Iterative Schemes to Solve Low-Dimensional Calibration Equations in Parallel MR Image Reconstruction with GRAPPA.

Omair Inam,Hammad Omer,Mahmood Qureshi,Shahzad A Malik

doi:10.1155/2017/3872783

Abstract

GRAPPA (Generalized Autocalibrating Partially Parallel Acquisition) is a widely used parallel MRI reconstruction technique. The processing of data from multichannel receiver coils may increase the storage and computational requirements of GRAPPA reconstruction. Random projection on GRAPPA (RP-GRAPPA) uses random projection (RP) method to overcome the computational overheads of solving large linear equations in the calibration phase of GRAPPA, saving reconstruction time. However, RP-GRAPPA compromises the reconstruction accuracy in case of large reductions in the dimensions of calibration equations. In this paper, we present the implementation of GRAPPA reconstruction method using potential iterative solvers to estimate the reconstruction coefficients from the randomly projected calibration equations. Experimental results show that the proposed methods withstand the reconstruction accuracy (visually and quantitatively) against large reductions in the dimension of linear equations, when compared with RP-GRAPPA reconstruction. Particularly, the proposed method using conjugate gradient for least squares (CGLS) demonstrates more savings in the computational time of GRAPPA, without significant loss in the reconstruction accuracy, when compared with RP-GRAPPA. It is also demonstrated that the proposed method using CGLS complements the channel compression method for reducing the computational complexities associated with higher channel count, thereby resulting in additional memory savings and speedup.

Highlights

Parallel imaging is an emerging technique to accelerate the MR data acquisition by undersampling the k-space data at each channel in multichannel coil arrays [1, 2]
We performed several experiments on in vivo datasets obtained using 8- and 12-channel receiver coils to validate the performance of the proposed methods (RP-conjugate gradient for least squares (CGLS)-GRAPPA and random projection (RP)-Heuristic rule-based gradient descent (HGD)-GRAPPA) in terms of reconstruction accuracy and computation time
We evaluated the proposed methods and RP-GRAPPA using 8-channel dataset to investigate the effect of reducing λ in the range between 4 and 1, on the computational time, reconstruction accuracy, and convergence behavior of CGLS and HGD algorithms

Summary

Introduction

Parallel imaging is an emerging technique to accelerate the MR data acquisition by undersampling the k-space data at each channel in multichannel coil arrays [1, 2]. In parallel imaging, undersampled data is acquired simultaneously by multiple channels and the image is reconstructed using parallel MRI (pMRI) techniques, for example, SENSE and GRAPPA [6, 7]. GRAPPA interpolates the undersampled k-space data of multichannel receiver coils by estimating the unknown reconstruction coefficients from the fully acquired autocalibration signal (ACS). Large computation and memory requirements due to higher channel count limit the efficiency and scalability of GRAPPA and other pMRI techniques on the computational platforms such as FPGAs and GPUs [9,10,11,12]

Objectives

Results

Conclusion