Title :
Multiplicative iterative algorithms for positive constrained reconstructions in emission and transmission tomography
Author_Institution :
Dept. of Stat., Macquarie Univ., North Ryde, NSW
Abstract :
This paper introduces a multiplicative iterative (MI) algorithm for image reconstructions in tomography. This algorithm can accommodate objective functions deduced from different probability models for measurements. Poisson and Gaussian (for both emission and transmission scans), or shifted Poisson (for precorrected PET and X-ray CT), are examples of such measurement probability models. This MI algorithm is very easy to implement and respects the positivity constraint. Furthermore, an exact or approximate line search step can be easily incorporated into this algorithm so that the objective functions are guaranteed to increase during the iterations.
Keywords :
Gaussian distribution; Poisson distribution; emission tomography; image reconstruction; iterative methods; Gaussian model; emission tomography; image reconstruction; measurement probability model; multiplicative iterative algorithm; objective function; positive constrained reconstruction; shifted Poisson model; transmission tomography; Computed tomography; Entropy; Image reconstruction; Independent component analysis; Iterative algorithms; Least squares approximation; Positron emission tomography; Probability; Statistics; X-ray imaging; EM; ISRA; Multiplicative iterative (MI) algorithm; emission and transmission tomography; line search;
Conference_Titel :
Biomedical Imaging: From Nano to Macro, 2008. ISBI 2008. 5th IEEE International Symposium on
Conference_Location :
Paris
Print_ISBN :
978-1-4244-2002-5
Electronic_ISBN :
978-1-4244-2003-2
DOI :
10.1109/ISBI.2008.4541177
