Compressed sensing based cone-beam computed tomography reconstruction with a first-order method

Kihwan Choi, Jing Wang, Lei Zhu, Tae Suk Suh, Stephen Boyd, Lei Xing

Research output: Contribution to journalArticle

167 Citations (Scopus)

Abstract

Purpose: This article considers the problem of reconstructing cone-beam computed tomography (CBCT) images from a set of undersampled and potentially noisy projection measurements. Methods: The authors cast the reconstruction as a compressed sensing problem based on l1 norm minimization constrained by statistically weighted least-squares of CBCT projection data. For accurate modeling, the noise characteristics of the CBCT projection data are used to determine the relative importance of each projection measurement. To solve the compressed sensing problem, the authors employ a method minimizing total-variation norm, satisfying a prespecified level of measurement consistency using a first-order method developed by Nesterov. Results: The method converges fast to the optimal solution without excessive memory requirement, thanks to the method of iterative forward and back-projections. The performance of the proposed algorithm is demonstrated through a series of digital and experimental phantom studies. It is found a that high quality CBCT image can be reconstructed from undersampled and potentially noisy projection data by using the proposed method. Both sparse sampling and decreasing x-ray tube current (i.e., noisy projection data) lead to the reduction of radiation dose in CBCT imaging. Conclusions: It is demonstrated that compressed sensing outperforms the traditional algorithm when dealing with sparse, and potentially noisy, CBCT projection views.

Original languageEnglish (US)
Pages (from-to)5113-5125
Number of pages13
JournalMedical Physics
Volume37
Issue number9
DOIs
StatePublished - Sep 2010

Fingerprint

Cone-Beam Computed Tomography
Least-Squares Analysis
Noise
X-Rays
Radiation

Keywords

  • compressed sensing
  • cone-beam computed tomography
  • Nesterov's first order method
  • weighted least-squares

ASJC Scopus subject areas

  • Biophysics
  • Radiology Nuclear Medicine and imaging
  • Medicine(all)

Cite this

Compressed sensing based cone-beam computed tomography reconstruction with a first-order method. / Choi, Kihwan; Wang, Jing; Zhu, Lei; Suh, Tae Suk; Boyd, Stephen; Xing, Lei.

In: Medical Physics, Vol. 37, No. 9, 09.2010, p. 5113-5125.

Research output: Contribution to journalArticle

Choi, Kihwan ; Wang, Jing ; Zhu, Lei ; Suh, Tae Suk ; Boyd, Stephen ; Xing, Lei. / Compressed sensing based cone-beam computed tomography reconstruction with a first-order method. In: Medical Physics. 2010 ; Vol. 37, No. 9. pp. 5113-5125.
@article{5b1ab95f5f1046a38edc84902fdb040f,
title = "Compressed sensing based cone-beam computed tomography reconstruction with a first-order method",
abstract = "Purpose: This article considers the problem of reconstructing cone-beam computed tomography (CBCT) images from a set of undersampled and potentially noisy projection measurements. Methods: The authors cast the reconstruction as a compressed sensing problem based on l1 norm minimization constrained by statistically weighted least-squares of CBCT projection data. For accurate modeling, the noise characteristics of the CBCT projection data are used to determine the relative importance of each projection measurement. To solve the compressed sensing problem, the authors employ a method minimizing total-variation norm, satisfying a prespecified level of measurement consistency using a first-order method developed by Nesterov. Results: The method converges fast to the optimal solution without excessive memory requirement, thanks to the method of iterative forward and back-projections. The performance of the proposed algorithm is demonstrated through a series of digital and experimental phantom studies. It is found a that high quality CBCT image can be reconstructed from undersampled and potentially noisy projection data by using the proposed method. Both sparse sampling and decreasing x-ray tube current (i.e., noisy projection data) lead to the reduction of radiation dose in CBCT imaging. Conclusions: It is demonstrated that compressed sensing outperforms the traditional algorithm when dealing with sparse, and potentially noisy, CBCT projection views.",
keywords = "compressed sensing, cone-beam computed tomography, Nesterov's first order method, weighted least-squares",
author = "Kihwan Choi and Jing Wang and Lei Zhu and Suh, {Tae Suk} and Stephen Boyd and Lei Xing",
year = "2010",
month = "9",
doi = "10.1118/1.3481510",
language = "English (US)",
volume = "37",
pages = "5113--5125",
journal = "Medical Physics",
issn = "0094-2405",
publisher = "AAPM - American Association of Physicists in Medicine",
number = "9",

}

TY - JOUR

T1 - Compressed sensing based cone-beam computed tomography reconstruction with a first-order method

AU - Choi, Kihwan

AU - Wang, Jing

AU - Zhu, Lei

AU - Suh, Tae Suk

AU - Boyd, Stephen

AU - Xing, Lei

PY - 2010/9

Y1 - 2010/9

N2 - Purpose: This article considers the problem of reconstructing cone-beam computed tomography (CBCT) images from a set of undersampled and potentially noisy projection measurements. Methods: The authors cast the reconstruction as a compressed sensing problem based on l1 norm minimization constrained by statistically weighted least-squares of CBCT projection data. For accurate modeling, the noise characteristics of the CBCT projection data are used to determine the relative importance of each projection measurement. To solve the compressed sensing problem, the authors employ a method minimizing total-variation norm, satisfying a prespecified level of measurement consistency using a first-order method developed by Nesterov. Results: The method converges fast to the optimal solution without excessive memory requirement, thanks to the method of iterative forward and back-projections. The performance of the proposed algorithm is demonstrated through a series of digital and experimental phantom studies. It is found a that high quality CBCT image can be reconstructed from undersampled and potentially noisy projection data by using the proposed method. Both sparse sampling and decreasing x-ray tube current (i.e., noisy projection data) lead to the reduction of radiation dose in CBCT imaging. Conclusions: It is demonstrated that compressed sensing outperforms the traditional algorithm when dealing with sparse, and potentially noisy, CBCT projection views.

AB - Purpose: This article considers the problem of reconstructing cone-beam computed tomography (CBCT) images from a set of undersampled and potentially noisy projection measurements. Methods: The authors cast the reconstruction as a compressed sensing problem based on l1 norm minimization constrained by statistically weighted least-squares of CBCT projection data. For accurate modeling, the noise characteristics of the CBCT projection data are used to determine the relative importance of each projection measurement. To solve the compressed sensing problem, the authors employ a method minimizing total-variation norm, satisfying a prespecified level of measurement consistency using a first-order method developed by Nesterov. Results: The method converges fast to the optimal solution without excessive memory requirement, thanks to the method of iterative forward and back-projections. The performance of the proposed algorithm is demonstrated through a series of digital and experimental phantom studies. It is found a that high quality CBCT image can be reconstructed from undersampled and potentially noisy projection data by using the proposed method. Both sparse sampling and decreasing x-ray tube current (i.e., noisy projection data) lead to the reduction of radiation dose in CBCT imaging. Conclusions: It is demonstrated that compressed sensing outperforms the traditional algorithm when dealing with sparse, and potentially noisy, CBCT projection views.

KW - compressed sensing

KW - cone-beam computed tomography

KW - Nesterov's first order method

KW - weighted least-squares

UR - http://www.scopus.com/inward/record.url?scp=77956331390&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77956331390&partnerID=8YFLogxK

U2 - 10.1118/1.3481510

DO - 10.1118/1.3481510

M3 - Article

VL - 37

SP - 5113

EP - 5125

JO - Medical Physics

JF - Medical Physics

SN - 0094-2405

IS - 9

ER -