Variations on Image Reconstruction:Weighted Back Projection and Fourier Expectation Maximization
- Submitting institution
-
Swansea University / Prifysgol Abertawe
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 21545
- Type
- E - Conference contribution
- DOI
-
10.2312/cgvc.20141201
- Title of conference / published proceedings
- Computer Graphics and Visual Computing
- First page
- 09
- Volume
- -
- Issue
- -
- ISSN
- -
- Open access status
- -
- Month of publication
- September
- Year of publication
- 2014
- URL
-
https://diglib.eg.org/handle/10.2312/cgvc.20141201.009-016
- Supplementary information
-
-
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
-
- Research group(s)
-
-
- Citation count
- -
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- Medical Imaging is crucial to society, and image quality of MRI, PET or CT Scan devices is paramount. Improving reconstruction from projections also allows reducing exposure to radiation. Our work merges the two mainstream reconstruction techniques, namely Filtered Backprojection (FBP) and Expectation Maximisation (EM). FBP is fast, as based on Fast Fourier Transforms while EM provides more accurate results but is generally slow and therefore underused. Our hybrid method is faster than regular EM reconstruction -by factor n/log(n)- while providing superior image reconstructions than FBP. Novelty aside, this work enlarges the scope of EM algorithms.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -