Total variation regularization for bioluminescence tomography with the split Bregman method

Jinchao Feng; Chenghu Qin; Kebin Jia; Shouping Zhu; Kai Liu; Dong Han; Xin Yang; Quansheng Gao; Jie Tian

doi:10.1364/AO.51.004501

Total variation regularization for bioluminescence tomography with the split Bregman method

Jinchao Feng, Chenghu Qin, Kebin Jia, Shouping Zhu, Kai Liu, Dong Han, Xin Yang, Quansheng Gao, and Jie Tian

Applied Optics
Vol. 51,
Issue 19,
pp. 4501-4512
(2012)
•https://doi.org/10.1364/AO.51.004501

Get Citation
Copy Citation Text
Jinchao Feng, Chenghu Qin, Kebin Jia, Shouping Zhu, Kai Liu, Dong Han, Xin Yang, Quansheng Gao, and Jie Tian, "Total variation regularization for bioluminescence tomography with the split Bregman method," Appl. Opt. 51, 4501-4512 (2012)

Export Citation
- BibTex
- Endnote (RIS)
- HTML
- Plain Text
Get PDF (1277 KB)
Set citation alerts
Save article

Related Topics
Table of Contents Category
- Medical Optics and Biotechnology
Optics & Photonics Topics
?

The topics in this list come from the OSA Optics and Photonics Topics applied to this article.
Previously assigned OCIS codes
- Image reconstruction techniques (170.3010)
- Spectroscopy, fluorescence and luminescence (170.6280)
- Tomography (170.6960)

About this Article
History
- Original Manuscript: July 28, 2011
- Revised Manuscript: May 10, 2012
- Manuscript Accepted: May 17, 2012
- Published: June 29, 2012
Virtual Issues
Virtual Journal for Biomedical Optics Vol. 7, Iss. 9

Not Accessible

Your account may give you access

Abstract

Regularization methods have been broadly applied to bioluminescence tomography (BLT) to obtain stable solutions, including $l_{2}$ and $l_{1}$ regularizations. However, $l_{2}$ regularization can oversmooth reconstructed images and $l_{1}$ regularization may sparsify the source distribution, which degrades image quality. In this paper, the use of total variation (TV) regularization in BLT is investigated. Since a nonnegativity constraint can lead to improved image quality, the nonnegative constraint should be considered in BLT. However, TV regularization with a nonnegativity constraint is extremely difficult to solve due to its nondifferentiability and nonlinearity. The aim of this work is to validate the split Bregman method to minimize the TV regularization problem with a nonnegativity constraint for BLT. The performance of split Bregman-resolved TV (SBRTV) based BLT reconstruction algorithm was verified with numerical and in vivo experiments. Experimental results demonstrate that the SBRTV regularization can provide better regularization quality over $l_{2}$ and $l_{1}$ regularizations.

Full Article | PDF Article

OSA Recommended Articles

Sparse reconstruction for quantitative bioluminescence tomography based on the incomplete variables truncated conjugate gradient method

Xiaowei He, Jimin Liang, Xiaorui Wang, Jingjing Yu, Xiaochao Qu, Xiaodong Wang, Yanbin Hou, Duofang Chen, Fang Liu, and Jie Tian
Opt. Express 18(24) 24825-24841 (2010)

Efficient sparse reconstruction algorithm for bioluminescence tomography based on duality and variable splitting

Wei Guo, Kebin Jia, Dong Han, Qian Zhang, Xueyan Liu, Jinchao Feng, Chenghu Qin, Xibo Ma, and Jie Tian
Appl. Opt. 51(23) 5676-5685 (2012)

Source Reconstruction for Spectrally-resolved Bioluminescence Tomography with Sparse A priori Information

Yujie Lu, Xiaoqun Zhang, Ali Douraghy, David Stout, Jie Tian, Tony F. Chan, and Arion F. Chatziioannou
Opt. Express 17(10) 8062-8080 (2009)

References

You do not have subscription access to this journal. Citation lists with outbound citation links are available to subscribers only. You may subscribe either as an OSA member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access OSA Member Subscription

Cited By

You do not have subscription access to this journal. Cited by links are available to subscribers only. You may subscribe either as an OSA member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access OSA Member Subscription

Figures (14)

You do not have subscription access to this journal. Figure files are available to subscribers only. You may subscribe either as an OSA member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access OSA Member Subscription

Tables (7)

You do not have subscription access to this journal. Article tables are available to subscribers only. You may subscribe either as an OSA member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access OSA Member Subscription

Equations (14)

You do not have subscription access to this journal. Equations are available to subscribers only. You may subscribe either as an OSA member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access OSA Member Subscription

Metrics

You do not have subscription access to this journal. Article level metrics are available to subscribers only. You may subscribe either as an OSA member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access OSA Member Subscription

Wavelength	600 nm	620 nm
$μ_{a}$ [ ${mm}^{- 1}$ ]	0.0281	0.0109
$μ_{s}^{'}$ [ ${mm}^{- 1}$ ]	1.6667	1.6129

Regularization method	Regularization parameter	Reconstructed position error	Reconstructed density error	MSE
$l_{2}$	$2.15 \times 10^{- 6}$	0.21 mm	77.81%	5.62
$l_{1}$	$4.91 \times 10^{- 5}$	0.89 mm	51.52%	6.39
SBRTV	$9.9 \times 10^{- 8}$	0.21 mm	6.67%	4.05

Regularization method	Regularization parameter	No. source	Reconstructed position error	Reconstructed density error	MSE
$l_{2}$	$4.56 \times 10^{- 5}$	No. 1	0.54 mm	83.77%	14.26
$l_{2}$	$4.56 \times 10^{- 5}$	No. 2	0.15 mm	86.44%	14.26
$l_{1}$	$2.36 \times 10^{- 4}$	No. 1	1.05 mm	39.35%	17.38
$l_{1}$	$2.36 \times 10^{- 4}$	No. 2	0.99 mm	46.26%	17.38
SBRTV	$1.15 \times 10^{- 7}$	No. 1	0.21 mm	25.35%	9.65
SBRTV	$1.15 \times 10^{- 7}$	No. 2	0.15 mm	48.38%	9.65

Regularization method	Regularization parameter	No. source	Reconstructed position error	Reconstructed density error	MSE
$l_{2}$	$4.56 \times 10^{- 5}$	No. 1	0.54 mm	85.81%	21.31
		No. 2	0.14 mm	84.5%
		No. 3	0.07 mm	67.91%
$l_{1}$	$2.36 \times 10^{- 4}$	No. 1	1.05 mm	44.74%	27.61
		No. 2	1.13 mm	67.24%
		No. 3	0.93 mm	13.36%
SBRTV	$1.15 \times 10^{- 7}$	No. 1	0.21 mm	49.65%	18.43
		No. 2	0.15 mm	45.03%
		No. 3	0.21 mm	5.39%

	650–700 nm		700–760 nm
Wavelength	$μ_{a}$	$μ_{s}^{'}$	$μ_{a}$	$μ_{s}^{'}$
Muscle	0.0771	0.4128	0.0396	0.3373
Heart	0.0521	0.9446	0.0282	0.8529
Lung	0.1741	2.1574	0.0899	2.0775
Liver	0.3129	0.6680	0.1646	0.6198
Bone	0.0540	2.4441	0.0280	2.2005

	$l_{2}$	$l_{1}$	SBRTV
Iteration number	19	22063	9
Time (seconds)	113	437	2084
$λ$	$4.7 \times 10^{- 5}$	$1.6 \times 10^{- 3}$	$4.7 \times 10^{- 5}$
Reconstructed error (mm)	1.66	0.84	0.84

	$l_{2}$	$l_{1}$	SBRTV
Iteration number	19	22063	9
Time (seconds)	113	437	2084
$λ$	$4.7 \times 10^{- 5}$	$1.6 \times 10^{- 3}$	$4.7 \times 10^{- 5}$
Reconstructed error (mm)	1.66	0.84	0.84

Abstract

References

Cited By

Figures (14)

Tables (7)

Equations (14)

Metrics

Login or Create Account