Paper Info
Title
Undersampled dynamic magnetic resonance imaging using kernel principal component analysis.
Abstract
Compressed sensing (CS) is a promising approach to accelerate dynamic magnetic resonance imaging (MRI). Most existing CS methods employ linear sparsifying transforms. The recent developments in non-linear or kernel-based sparse representations have been shown to outperform the linear transforms. In this paper, we present an iterative non-linear CS dynamic MRI reconstruction framework that uses the kernel principal component analysis (KPCA) to exploit the sparseness of the dynamic image sequence in the feature space. Specifically, we apply KPCA to represent the temporal profiles of each spatial location and reconstruct the images through a modified pre-image problem. The underlying optimization algorithm is based on variable splitting and fixed-point iteration method. Simulation results show that the proposed method outperforms conventional CS method in terms of aliasing artifact reduction and kinetic information preservation.
Year
DOI
Venue
2014
10.1109/EMBC.2014.6943894
EMBC
Keywords
DocType
Volume
modified preimage problem,temporal profiles,mri,kinetic information preservation,image reconstruction,biomedical mri,image sampling,nonlinear based sparse representations,compressed sensing,feature extraction,variable splitting,image sequences,kernel-based sparse representations,aliasing artifact reduction,dynamic image sequence,linear sparsifying transforms,undersampled dynamic magnetic resonance imaging,feature space,underlying optimization algorithm,fixed-point iteration method,principal component analysis,kernel principal component analysis,kpca,iterative methods,iterative nonlinear cs dynamic mri reconstruction framework,medical image processing
Conference
2014
ISSN
Citations 
PageRank 
1557-170X
4
0.45
References 
Authors
9
2
Name
Order
Citations
PageRank
Yanhua Wang1476.35
Leslie Ying224029.08