Concurrent Iterative Reconstruction Algorithms use projection data in the iterative process as the data become available during the SPECT acquisition process and continue iterations in the post-acquisition period as conventional iterative algorithms. Because projections acquired early are processed more than later projections, regional inhomogeneities may exist in the initial image estimates but decrease with further post-acquisition iteration. Regularization done either during the acquisition or post-acquisition iterations further reduces regional inhomogeneities. The authors tested statistical differences in regions throughout the reconstructed image to determine the minimal number of post-acquisition iterations and type of regularization needed to reach an image that is inter-regionally consistent. The algorithms provide images free of reconstruction inhomogeneities and can offer a reduction in post-acquisition reconstruction time when compared to conventional iterative algorithms.
[1]
B. Hasegawa,et al.
Iterative concurrent reconstruction algorithms for emission computed tomography.
,
1994,
Physics in medicine and biology.
[2]
Bruce H. Hasegawa,et al.
The simulation of concurrent iterative reconstruction algorithms for single photon emission computed tomography
,
1994,
Proceedings of 1994 IEEE Nuclear Science Symposium - NSS'94.
[3]
H. Malcolm Hudson,et al.
Accelerated image reconstruction using ordered subsets of projection data
,
1994,
IEEE Trans. Medical Imaging.
[4]
L. Shepp,et al.
Maximum Likelihood Reconstruction for Emission Tomography
,
1983,
IEEE Transactions on Medical Imaging.
[5]
L. Shepp,et al.
A Statistical Model for Positron Emission Tomography
,
1985
.