Object-based 3D binary reconstruction from sparse projections in cone beam CT: Comparison of three projection operators

We present herein a level set approach to the X-ray tomography problem with sparse projection data and study the impact of the projection operator on the binary reconstruction accuracy and computation time. The comparison is carried out on three projectors: the Separable Footprint (Trapeze-Trapeze, SF-TT) [3], a classical Raydriven (RD) and a Simplified version of the Distance-Driven (SDD) projector respectively. The performance, are evaluated for each operator, on a binary 3D Shepp-Logan phantom by varying the number of projections from 5 to 13, and considering noise free and noisy cone beam projection data.

Keywords

3D tomographic reconstruction Projection matrix

Domains

Bioengineering Signal and Image processing Signal and Image Processing

Fichier principal

ISBI_2013_Oukili_Chen_Nunes_Huo_Toumoulin_Final_Copy.pdf (453.81 Ko)

Origin : Files produced by the author(s)

Christine Toumoulin : Connect in order to contact the contributor

https://www.hal.inserm.fr/inserm-00874956

Submitted on : Saturday, October 19, 2013-7:38:18 PM

Last modification on : Thursday, March 16, 2023-3:10:09 PM

Long-term archiving on: Monday, January 20, 2014-4:25:44 AM

Dates and versions

inserm-00874956 , version 1 (19-10-2013)

Identifiers

HAL Id : inserm-00874956 , version 1
DOI : 10.1109/ISBI.2013.6556764

Cite

Ahmed Oukili, Jean-Claude Nunes, Chen Yang, Limin M. Luo, Christine Toumoulin. Object-based 3D binary reconstruction from sparse projections in cone beam CT: Comparison of three projection operators. International Symposium on BIOMEDICAL IMAGING: From Nano to Macro, Apr 2013, San Francisco, United States. pp.1276 - 1279, ⟨10.1109/ISBI.2013.6556764⟩. ⟨inserm-00874956⟩

Export

BibTeX TEI Dublin Core DC Terms EndNote Datacite

Collections

INSERM UNIV-RENNES1 LTSI CRIBS UR1-MATH-STIC UNIV-RENNES UR1-MATH-NUM UR1-BIO-SA

527 View

359 Download