Paper Abstract and Keywords |
Presentation |
2016-05-19 15:10
High accuracy reconstruction algorithm for CS-MRI using SDMM Motoi Shibata, Norihito Inamuro, Takashi Ijiri, Akira Hirabayashi (Ritsumeikan Univ.) SIP2016-12 IE2016-12 PRMU2016-12 MI2016-12 |
Abstract |
(in Japanese) |
(See Japanese page) |
(in English) |
We propose a high accuracy magnetic resonance imaging (MRI) reconstruction algorithm from compressively sampled measurements using a convex optimization technique. Lustig et al. proposed the compressed sensing MRI (CS-MRI) technique, in which MR images are reconstructed by minimizing a cost function defined by the sum of the data fidelity term, the l1-norm of sparsifying transform coefficients, and a total-variation (TV). Since the absolute values in both l1-norm and TV are not differentiable at the origin, they approximated it by adding a small positive constant in the square root. Then, a nonlinear conjugate gradient descent algorithm was exploited to minimize the approximated cost function. The obtained solution is also an approximated one, thus of low-quality. Hence, in this paper, we propose an algorithm that obtains a rigorous solution to the minimization problem without any approximation based on the simultaneous direction method of multipliers (SDMM), one of the convex optimization techniques. A simple application of SDMM to CS-MRI can not be implemented on computers because of the matrix size that is proportional to the square of the image size. We solve this problem using eigen value decompositions. Simulations using real MR images show that the proposed algorithm outperforms the conventional one irrespective of compression ratio and random sensing scenarios. |
Keyword |
(in Japanese) |
(See Japanese page) |
(in English) |
MRI / compressed sensing / total-variation / convex optimization / ADMM / SDMM / / |
Reference Info. |
IEICE Tech. Rep., vol. 116, no. 36, SIP2016-12, pp. 59-64, May 2016. |
Paper # |
SIP2016-12 |
Date of Issue |
2016-05-12 (SIP, IE, PRMU, MI) |
ISSN |
Print edition: ISSN 0913-5685 Online edition: ISSN 2432-6380 |
Copyright and reproduction |
All rights are reserved and no part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Notwithstanding, instructors are permitted to photocopy isolated articles for noncommercial classroom use without fee. (License No.: 10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
Download PDF |
SIP2016-12 IE2016-12 PRMU2016-12 MI2016-12 |
Conference Information |
Committee |
PRMU IE MI SIP |
Conference Date |
2016-05-19 - 2016-05-20 |
Place (in Japanese) |
(See Japanese page) |
Place (in English) |
|
Topics (in Japanese) |
(See Japanese page) |
Topics (in English) |
|
Paper Information |
Registration To |
SIP |
Conference Code |
2016-05-PRMU-IE-MI-SIP |
Language |
Japanese |
Title (in Japanese) |
(See Japanese page) |
Sub Title (in Japanese) |
(See Japanese page) |
Title (in English) |
High accuracy reconstruction algorithm for CS-MRI using SDMM |
Sub Title (in English) |
|
Keyword(1) |
MRI |
Keyword(2) |
compressed sensing |
Keyword(3) |
total-variation |
Keyword(4) |
convex optimization |
Keyword(5) |
ADMM |
Keyword(6) |
SDMM |
Keyword(7) |
|
Keyword(8) |
|
1st Author's Name |
Motoi Shibata |
1st Author's Affiliation |
Ritsumeikan University (Ritsumeikan Univ.) |
2nd Author's Name |
Norihito Inamuro |
2nd Author's Affiliation |
Ritsumeikan University (Ritsumeikan Univ.) |
3rd Author's Name |
Takashi Ijiri |
3rd Author's Affiliation |
Ritsumeikan University (Ritsumeikan Univ.) |
4th Author's Name |
Akira Hirabayashi |
4th Author's Affiliation |
Ritsumeikan University (Ritsumeikan Univ.) |
5th Author's Name |
|
5th Author's Affiliation |
() |
6th Author's Name |
|
6th Author's Affiliation |
() |
7th Author's Name |
|
7th Author's Affiliation |
() |
8th Author's Name |
|
8th Author's Affiliation |
() |
9th Author's Name |
|
9th Author's Affiliation |
() |
10th Author's Name |
|
10th Author's Affiliation |
() |
11th Author's Name |
|
11th Author's Affiliation |
() |
12th Author's Name |
|
12th Author's Affiliation |
() |
13th Author's Name |
|
13th Author's Affiliation |
() |
14th Author's Name |
|
14th Author's Affiliation |
() |
15th Author's Name |
|
15th Author's Affiliation |
() |
16th Author's Name |
|
16th Author's Affiliation |
() |
17th Author's Name |
|
17th Author's Affiliation |
() |
18th Author's Name |
|
18th Author's Affiliation |
() |
19th Author's Name |
|
19th Author's Affiliation |
() |
20th Author's Name |
|
20th Author's Affiliation |
() |
Speaker |
Author-1 |
Date Time |
2016-05-19 15:10:00 |
Presentation Time |
30 minutes |
Registration for |
SIP |
Paper # |
SIP2016-12, IE2016-12, PRMU2016-12, MI2016-12 |
Volume (vol) |
vol.116 |
Number (no) |
no.36(SIP), no.37(IE), no.38(PRMU), no.39(MI) |
Page |
pp.59-64 |
#Pages |
6 |
Date of Issue |
2016-05-12 (SIP, IE, PRMU, MI) |
|