2014
DOI: 10.1109/tns.2013.2283529
|View full text |Cite
|
Sign up to set email alerts
|

Superiorization of the ML-EM Algorithm

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
43
0

Year Published

2014
2014
2021
2021

Publication Types

Select...
7
2

Relationship

0
9

Authors

Journals

citations
Cited by 36 publications
(43 citation statements)
references
References 34 publications
0
43
0
Order By: Relevance
“…Superiorization work with other target functions such as total variation (TV) appears in, e.g., [13, 15, 27]. Superiorization work on Case 3, where M is the solution set of a maximum likelihood optimization problem appears in [23, 28, 30]. …”
Section: The Superiorization Methodologymentioning
confidence: 99%
“…Superiorization work with other target functions such as total variation (TV) appears in, e.g., [13, 15, 27]. Superiorization work on Case 3, where M is the solution set of a maximum likelihood optimization problem appears in [23, 28, 30]. …”
Section: The Superiorization Methodologymentioning
confidence: 99%
“…We follow the same algorithm flowchart as in Censor et al (2014) and Garduno and Herman (2014). The superiorized SART algorithm is in table 1.…”
Section: Methodsmentioning
confidence: 99%
“…Specifically, the basic algorithm that solves the data fidelity term remains the same but the intermediate solution is steered towards the direction to reduce the penalty term. Several studies were performed to superiorize the ML-EM algorithm for emission reconstruction (Garduno and Herman 2014, Luo and Zhou 2014). TV and the l 1 -norm of a wavelet transform were implemented as the regularizers.…”
Section: Introductionmentioning
confidence: 99%
“…In the first of these, the saem [7] is compared with the em [15] considering two kind of superiorization sequences for non-negatively constrained Total Variation as a secondary criteria. The first superiorization sequence is produced according to Algorithm 2, following the ideas presented in [5], the second superiorization sequence is produced by the fgp algorithm from [1].…”
Section: Numerical Experimentationmentioning
confidence: 99%