A novel dual single photon emission computerized tomography (SPECT) imaging system equipped with a modular coded aperture and a conventional parallel hole collimator is introduced An expectation maximization type blind deconvolution algorithm (EMBD) based on maximum likelihood estimation is derived for restoration of SPECT planar projections acquired by the dual imaging system. The uniformly redundant array (URA) technique is used for decoding the shadowgram acquired by the coded aperture camera head. Results of real phantom data are presented. The new configuration of file dual SPECT imaging system considerably improves the count sensitivity while still maintaining (or slightly improving) the conventional resolutions of high resolution collimation. The results also demonstrate the feasibility of the EMBD algorithm in improving image resolution and reducing background noise and non-uniformity artifacts of the decoded images. Thus, the new imaging system associated with the image restoration algorithm may be an advance of low-count nuclear imaging.
CITATION STYLE
Liu, Y. H., Rangarajan, A., Gagnon, D., Therrien, M., Sinusas, A. J., Wackers, F. J. T., & Zubal, G. (1997). New methods for SPECT imaging and pre-reconstruction restoration. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 1230, pp. 531–536). Springer Verlag. https://doi.org/10.1007/3-540-63046-5_55
Mendeley helps you to discover research relevant for your work.