Please use this identifier to cite or link to this item: https://scholarhub.balamand.edu.lb/handle/uob/1177
Title: Numerical results for iterative image reconstruction in half-scan and non-uniform attenuated SPECT
Authors: Nasr, Elie
Greige, Hanna 
Affiliations: Department of Mathematics 
Keywords: Half-scan
Single-photon emission computed tomography
Algebraic reconstruction technique
Attenuation correction
Issue Date: 2011
Conference: Conference on Pure, Applied, and Computational Mathematics (April 2011 : LAU, Beirut Lebanon) 
Abstract: 
The quality of reconstructed image in Single-Photon Emission Computed Tomography (SPECT) is strongly degraded by the photon attenuation. In general, the attenuation correction on SPECT images requires the data to be known over 2π (full-scan). The reduction of data acquisition from 2π to π (half-scan) in SPECT is recommended because it reduces the scanning time; thereby, minimizing the patients motion and making the exam less uncomfortable for the patient. Algorithms and numerical research has been developed on image reconstruction from data acquired over π in SPECT with non-uniform attenuation. However, it remains theoretically unknown whether data, acquired only over π in SPECT with nonuniform attenuation, contain complete information for accurate image reconstruction. In this work, we present numerical results on image reconstruction in halfscan SPECT with non-uniform attenuation. The numerical simulations are based on a new iterative reconstruction algorithm that introduces exact and implicit attenuation correction. Specifically, we show that the algorithm is able to reconstruct images, in half-scan SPECT, with quality very similar to those reconstructed in full-scan SPECT. Moreover, the numerical simulations show that the algorithm is stable and convergent.
URI: https://scholarhub.balamand.edu.lb/handle/uob/1177
Type: Conference Paper
Appears in Collections:Department of Mathematics

Show full item record

Record view(s)

44
checked on Nov 20, 2024

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.