scholarly journals High-resolution dynamic speech imaging with joint low-rank and sparsity constraints

2014 ◽  
Vol 73 (5) ◽  
pp. 1820-1832 ◽  
Author(s):  
Maojing Fu ◽  
Bo Zhao ◽  
Christopher Carignan ◽  
Ryan K. Shosted ◽  
Jamie L. Perry ◽  
...  
Keyword(s):  
High Resolution ◽  
Low Rank ◽  
Sparsity Constraints

scholarly journals High-Resolution Cardiovascular MRI by Integrating Parallel Imaging With Low-Rank and Sparse Modeling

2013 ◽  
Vol 60 (11) ◽  
pp. 3083-3092 ◽  
Author(s):  
Anthony G. Christodoulou ◽  
Haosen Zhang ◽  
Bo Zhao ◽  
T. Kevin Hitchens ◽  
Chien Ho ◽  
...  
Keyword(s):  
High Resolution ◽  
Parallel Imaging ◽  
Low Rank ◽  
Cardiovascular Mri ◽  
Sparse Modeling

scholarly journals Exploring Inter-Instance Relationships within the Query Set for Robust Image Set Matching

Sensors ◽  
2019 ◽  
Vol 19 (22) ◽  
pp. 5051 ◽  
Author(s):  
Deyin Liu ◽  
Chengwu Liang ◽  
Zhiming Zhang ◽  
Lin Qi ◽  
Brian C. Lovell
Keyword(s):  
Optimization Problems ◽  
Low Rank ◽  
Collaborative Representation ◽  
Alternating Direction ◽  
Proposed Model ◽  
Class Level ◽  
Sparsity Constraints ◽  
Image Set ◽  
Robust Image ◽  
Public Datasets

Image set matching (ISM) has attracted increasing attention in the field of computer vision and pattern recognition. Some studies attempt to model query and gallery sets under a joint or collaborative representation framework, achieving impressive performance. However, existing models consider only the competition and collaboration among gallery sets, neglecting the inter-instance relationships within the query set which are also regarded as one important clue for ISM. In this paper, inter-instance relationships within the query set are explored for robust image set matching. Specifically, we propose to represent the query set instances jointly via a combined dictionary learned from the gallery sets. To explore the commonality and variations within the query set simultaneously to benefit the matching, both low rank and class-level sparsity constraints are imposed on the representation coefficients. Then, to deal with nonlinear data in real scenarios, the`kernelized version is also proposed. Moreover, to tackle the gross corruptions mixed in the query set, the proposed model is extended for robust ISM. The optimization problems are solved efficiently by employing singular value thresholding and block soft thresholding operators in an alternating direction manner. Experiments on five public datasets demonstrate the effectiveness of the proposed method, comparing favorably with state-of-the-art methods.

Efficient parallel reconstruction for high resolution multishot spiral diffusion data with low rank constraint

2016 ◽  
Vol 77 (3) ◽  
pp. 1359-1366 ◽  
Author(s):  
Congyu Liao ◽  
Ying Chen ◽  
Xiaozhi Cao ◽  
Song Chen ◽  
Hongjian He ◽  
...  
Keyword(s):  
High Resolution ◽  
Low Rank ◽  
Diffusion Data ◽  
Rank Constraint ◽  
Parallel Reconstruction

scholarly journals Reconstruction of Undersampled Big Dynamic MRI Data Using Non-Convex Low-Rank and Sparsity Constraints

Sensors ◽  
2017 ◽  
Vol 17 (3) ◽  
pp. 509 ◽  
Author(s):  
Ryan Liu ◽  
Lin Shi ◽  
Simon Yu ◽  
Naixue Xiong ◽  
Defeng Wang
Keyword(s):  
Dynamic Mri ◽  
Low Rank ◽  
Sparsity Constraints

scholarly journals Achieving High-Resolution Whole-Brain Slab 1H-MRSI with Compressed-Sensing and Low-Rank Reconstruction at 7 Tesla

2020 ◽  
Author(s):  
Antoine Klauser ◽  
Bernhard Strasser ◽  
Bijaya Thapa ◽  
Francois Lazeyras ◽  
Ovidiu Andronesi
Keyword(s):  
High Resolution ◽  
Compressed Sensing ◽  
Three Dimensions ◽  
Low Rank ◽  
7 Tesla ◽  
Brain Anatomy ◽  
Ultra High Field ◽  
Whole Brain ◽  
Low Sensitivity ◽  
Generalized Variation

Low sensitivity MR techniques such as magnetic resonance spectroscopic imaging (MRSI) greatly benefit from the gain in signal-to-noise (SNR) provided by ultra-high field MR. High-resolution and whole-brain slab MRSI remains however very challenging due to lengthy acquisition, low signal, lipid contamination and field inhomogeneity. In this study, we propose an acquisition-reconstruction scheme that combines a 1H-FID-MRSI sequence with compressed sensing acceleration and low-rank modeling with total-generalized-variation constraint to achieve metabolite imaging in two and three dimensions at 7 Tesla. The resulting images and volumes reveal highly detailed distributions that are specific to each metabolite and follow the underlying brain anatomy. The MRSI method was validated in a high-resolution phantom containing fine metabolite structures, and in 3 healthy volunteers. This new application of compressed sensing acceleration paves the way for high-resolution MRSI in clinical setting with acquisition times of 5 min for 2D MRSI at 2.5 mm and of 20 min for 3D MRSI at 3.3 mm isotropic.

scholarly journals Whole-Brain High-Resolution Metabolite Mapping with 3D Compressed-Sensing-SENSE-LowRank 1H FID-MRSI

2020 ◽  
Author(s):  
Antoine Klauser ◽  
Paul Klauser ◽  
Frédéric Grouiller ◽  
Sebastien Courvoisier ◽  
Francois Lazeyras
Keyword(s):  
High Resolution ◽  
Brain Regions ◽  
Low Rank ◽  
Acquisition Time ◽  
Resonance Spectroscopy ◽  
Brain Anatomy ◽  
Reconstruction Technique ◽  
Whole Brain ◽  
Isotropic Resolution

There is a growing interest of the neuroscience community to map the distribution of brain metabolites in vivo. Magnetic resonance spectroscopy imaging (MRSI) is often limited by either a poor spatial resolution and/or a long acquisition time which severely limits its applications for clinical or research purposes. We developed a novel acquisition-reconstruction technique combining fast 1H-FID-MRSI sequence accelerated by random k-space undersampling and a low-rank and total-generalized variation (TGV) constrained model. This framework was applied to the brain of four healthy volunteers. Following 20 min acquisition, reconstruction and quantification, the resulting metabolic maps with a 5 mm isotropic resolution reflected the detailed neurochemical composition of all brain regions and revealed part of the underlying brain anatomy. Contrasts and features from the 3D metabolite distributions were in agreement with the literature and consistent across the four subjects. The successful combination of the 3D 1H-FID-MRSI with a constrained reconstruction enables the detailed mapping of metabolite concentrations at high-resolution in the whole brain and with an acquisition time that is compatible with clinical or research settings.

scholarly journals Fast high-resolution brain metabolite mapping on a clinical 3T MRI by accelerated 1H-FID-MRSI and low-rank constrained reconstruction

2018 ◽  
Vol 81 (5) ◽  
pp. 2841-2857 ◽  
Author(s):  
Antoine Klauser ◽  
Sebastien Courvoisier ◽  
Jeffrey Kasten ◽  
Michel Kocher ◽  
Matthieu Guerquin-Kern ◽  
...  
Keyword(s):  
High Resolution ◽  
Low Rank ◽  
Constrained Reconstruction ◽  
3T Mri ◽  
Brain Metabolite
Sign in / Sign up

Export Citation Format

Share Document