Iterative Image Reconstruction Methods applied to data from a prototype small-animal PET

The purpose of this study is to evaluate the average performance of algebraic and statistical iterative reconstruction methods, using phantom data from a prototype small-animal PET system. The algorithms that are being compared are the simultaneous versions of ART (SART) and MART (SMART), EM-ML, ISRA and WLS. The evaluation study was based on reconstructed image quality, as it is derived from visual inspection, normalized profiles, cross-correlation coefficient and CNRs (contrast-to-noise ratios) of specific ROIs (region-of-interest). In general EM-ML and ISRA present similar reconstruction time and minor differences in reconstructed image quality. Slightly superior performances show WLS and SART while SMART is not adequate for reconstruction of PET data.