Effective emission tomography image reconstruction algorithms for SPECT data

J. Ramírez, J. M. Górriz, M. Gómez-Río, A. Romero, R. Chaves, A. Lassl, A. Rodríguez, C. G. Puntonet, F. Theis, E. Lang

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

41 Scopus citations

Abstract

Medical image reconstruction from projections is computationally intensive task that demands solutions for reducing the processing delay in clinical diagnosis applications. This paper analyzes reconstruction methods combined with pre- and post-filtering for Single Photon Emission Computed Tomography (SPECT) in terms of convergence speed and image quality. The evaluation is performed by means of an image database taken from a concurrent study investigating the use of SPECT as a diagnostic tool for the early onset of Alzheimer-type dementia. Filtered backprojection (FBP) methods combined with frequency sampling 2D pre- and post-filtering provides a good trade-off between image quality and delay. Maximum likelihood expectation maximization (ML-EM) improves the quality of the reconstructed image but with a considerable increase in processing delay. To overcome this problem the ordered subsets expectation maximization (OS-EM) method is found to be an effective algorithm for reducing the computational cost with an image quality similar to ML-EM.

Original languageEnglish
Title of host publicationComputational Science - ICCS 2008 - 8th International Conference, Proceedings
Pages741-748
Number of pages8
EditionPART 1
DOIs
StatePublished - 2008
Externally publishedYes
Event8th International Conference on Computational Science, ICCS 2008 - Krakow, Poland
Duration: 23 Jun 200825 Jun 2008

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
NumberPART 1
Volume5101 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference8th International Conference on Computational Science, ICCS 2008
Country/TerritoryPoland
CityKrakow
Period23/06/0825/06/08

Fingerprint

Dive into the research topics of 'Effective emission tomography image reconstruction algorithms for SPECT data'. Together they form a unique fingerprint.

Cite this