2020
DOI: 10.1016/j.bspc.2020.102095
|View full text |Cite
|
Sign up to set email alerts
|

A novel reconstruction approach combining OSEM and split Bregman method for low dose CT

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
6
0

Year Published

2022
2022
2024
2024

Publication Types

Select...
6

Relationship

1
5

Authors

Journals

citations
Cited by 7 publications
(6 citation statements)
references
References 10 publications
0
6
0
Order By: Relevance
“…As mentioned earlier, the DCT signal is collected from a total of 48 S‐D measurements, which is far less than the tissue elements (i.e., 1536 unknown αD B ), and results in a severely ill‐posed problem of Equation (). According to our previous efforts, the split‐Bregman algorithm combined with total variation (TV) regularization, namely, Bregman‐TV algorithm, would greatly improve the quality of image reconstruction 44,45 . Specifically, let v = αD B , b = Sl , the target function of the Bregman‐TV proposed for NL‐DCT is as follow. v*=argmin‖‖vitalicTV+μ2‖‖italicBvgoodbreak−b0.1em22s.t.vi0i=1,2,,n$$ {\displaystyle \begin{array}{c}{v}^{\ast }=\mathrm{argmin}{\left\Vert v\right\Vert}_{TV}\kern0.3em +\kern1em \frac{\mu }{2}{\left\Vert Bv-b\kern0.1em \right\Vert}_2^2\kern0.8000001em \\ {}\mathrm{s}.\mathrm{t}.\kern1.75em {v}_i\ge 0\left(i=1,\kern0.4em 2,\kern0.4em \dots \kern0.4em ,n\right)\end{array}} $$ here, non‐negativity of solutions is enforced.…”
Section: Methodsmentioning
confidence: 99%
See 1 more Smart Citation
“…As mentioned earlier, the DCT signal is collected from a total of 48 S‐D measurements, which is far less than the tissue elements (i.e., 1536 unknown αD B ), and results in a severely ill‐posed problem of Equation (). According to our previous efforts, the split‐Bregman algorithm combined with total variation (TV) regularization, namely, Bregman‐TV algorithm, would greatly improve the quality of image reconstruction 44,45 . Specifically, let v = αD B , b = Sl , the target function of the Bregman‐TV proposed for NL‐DCT is as follow. v*=argmin‖‖vitalicTV+μ2‖‖italicBvgoodbreak−b0.1em22s.t.vi0i=1,2,,n$$ {\displaystyle \begin{array}{c}{v}^{\ast }=\mathrm{argmin}{\left\Vert v\right\Vert}_{TV}\kern0.3em +\kern1em \frac{\mu }{2}{\left\Vert Bv-b\kern0.1em \right\Vert}_2^2\kern0.8000001em \\ {}\mathrm{s}.\mathrm{t}.\kern1.75em {v}_i\ge 0\left(i=1,\kern0.4em 2,\kern0.4em \dots \kern0.4em ,n\right)\end{array}} $$ here, non‐negativity of solutions is enforced.…”
Section: Methodsmentioning
confidence: 99%
“…According to our previous efforts, the split-Bregman algorithm combined with total variation (TV) regularization, namely, Bregman-TV algorithm, would greatly improve the quality of image reconstruction. 44,45 Specifically, let v = αD B , b = Sl, the target function of the Bregman-TV proposed for NL-DCT is as follow.…”
Section: Image Reconstruction Approachmentioning
confidence: 99%
“…( 6) and applying the optimality conditions, the deformation leads to the basic Bregman iterative model as shown in eq. ( 8) [27],…”
Section: Rof-tvmentioning
confidence: 99%
“…[ 11,12 ] In contrast, the iterative algorithm uses a system model and regularization, which makes the quality of reconstructed images better than the FBP algorithm, but the computational complexity of each iteration process in the iterative algorithm is equivalent to one FBP algorithm, so the reconstruction speed of the iterative algorithm is usually much slower than that of the FBP algorithm. [ 13,14 ]…”
Section: Introductionmentioning
confidence: 99%
“…[11,12] In contrast, the iterative algorithm uses a system model and regularization, which makes the quality of reconstructed images better than the FBP algorithm, but the computational complexity of each iteration process in the iterative algorithm is equivalent to one FBP algorithm, so the reconstruction speed of the iterative algorithm is usually much slower than that of the FBP algorithm. [13,14] In recent years, deep learning has been applied and made significant progress in various fields such as image restoration, image segmentation, and object detection. Many researchers have also applied deep learning techniques to PET reconstruction, including deep learning as a pre-processing tool in the reconstruction process for processing projection domain data.…”
Section: Introductionmentioning
confidence: 99%