Distance Driven Projection and Backprojection for Spherically Symmetric Basis Functions

Yulia Levakhina, Thorsten M. Buzug

Abstract

Forward- and backprojecton pair plays an important role in computed tomography (CT). Since they are used in clinical routine for filtered backprojection (FBP) reconstruction, in iterative reconstruction methods, for artifact correction and simulation purposes, they have to be fast, accurate and memory efficient. Recently, a distance-driven approach for pixel basis functions has been proposed. At the moment, it is a state-of-the-art method that is almost artifact free, fast and has a predictable memory pattern access. In the work presented here, the distance-driven approach for the two-dimensional case is extended for spherically symmetric Kaiser-Bessel basis functions. Usage of these basis functions allows for constructing a smooth and continuous function for image representation. A circle and Shepp-Logan simulated software phantom as well as an abdomen phantom, scanned by a multislice CT scanner with flying-focal spot (Somatom Definition AS; Siemens Medical Solutions) have been used to evaluate the presented forward- and backprojecton pair. Images have been reconstructed using FBP and iterative reconstruction with the distance-driven approach for pixel and spherically symmetric basis functions and compared in terms of image quality. The obtained quantitative results show the potential of the proposed method.

Original languageEnglish
Title of host publicationIEEE Nuclear Science Symposuim & Medical Imaging Conference
Number of pages4
PublisherIEEE
Publication date01.12.2010
Pages2894-2897
Article number5874325
ISBN (Print)978-1-4244-9106-3 , 978-1-4244-9104-9
ISBN (Electronic)978-1-4244-9105-6
DOIs
Publication statusPublished - 01.12.2010
Event2010 IEEE Nuclear Science Symposium, Medical Imaging Conference - Knoxville, United States
Duration: 30.10.201006.11.2010

Fingerprint

Dive into the research topics of 'Distance Driven Projection and Backprojection for Spherically Symmetric Basis Functions'. Together they form a unique fingerprint.

Cite this