Jafar Zamani

MSc.

Almuni: Now works at Bonab University
Personal Homepage: Not Available
Email: zamani.jafar{at}yahoo{dot}com
Supervisors: Dr. Abbas Nasiraei Moghaddam, Dr. Hamidreza Saligherad

EDUCATION

Research Subject

Accelerated MR Angiography through decreasing the phase encoding steps: Comparison of Parallel acquisition and Compressed Sensing

Abstract

The MRI speed is slow. Since MRI data is collected in K-space (or frequency domain) through a number of phase encoding steps, the acquisition time is proportional to this number and therefore can be decreased by under-sampling of phase encoding lines. Compressed Sensing (CS) and parallel MRI (pMRI) are methods to reduce MRI’s scan time. Compressed Sensing (CS) is a sampling theory with potential to reconstruct sparse images from a small number of randomly sampled data. There are three necessary conditions for CS as follows: 1) sparsity, 2) incoherent under-sampling, and 3) non-linear reconstruction. Parallel MR imaging (pMRI) employs multiple receiver coils and simultaneously acquires multiple data measurements from the coils, each weighted by a distinctive coil sensitivity profile. The primary drawback that limit clinical application these methods is requirement of accurate knowledge of sensitivity information of coils at pixel in image. Due to noise and spin density variation determine of coil sensitivity information is difficult. In this study, we focus on three necessary conditions of CS. we developed a fast, accurate CS-based algorithm for reconstruction of diagnostic contrast-enhanced MRA. This algorithm exploits Split Bregman method to iteratively minimize the objective function which is the sum of error and sparsity. In this study, we proposed the Principle Component Analysis (PCA) and Singular Value Decomposition (SVD) for weighting the sparsity in the CS formulation. Considering the dimension reduction property of PCA and SVD, it is suitable for weighting the sparse transform in the CS algorithm. In this project, we compared the efficiency of three different K-space under-sampling schemes; all patterns fully select the vicinity of the K-space centre, while they sample other K-space regions with different probability density functions. Cardiac MRI cine benefits from both temporal and spatial sparsity. In this work, we introduced a CS-based method for reconstruction of time-varying K-space data by exploiting spatio-temporal sparsity of cardiac MRI images. We proposed a new method to increase under-sampling rate and to expedite reconstruction time in CS theory. In this study, we reconstructed eight (one in every three) frames through CS using Gradient Projection for Sparse Reconstruction (GPSR) algorithm. The remaining 15 frames were reconstructed through a combination of CS and temporal information (TI). Sampling rate for the CS and CS-TI slices was set to 0.5 and 0.3, respectively. For further accelerating MRI acquisition time, we propose a method to combine two method for combine pMRI and CS. We combine CS and PILS, CS and SENSE, that employs CS at the first step to reconstruct a set of aliased reduced-field-of-view (FOV) images in each channel, and then apply PILS/SENSE to reconstruct the final image.

Publications

Copyright © Advanced Medical Imaging Research Laboratory. All Rights Reserved.