An evaluation of exact and approximate 3-D reconstruction algorithms for ahigh-resolution, small-animal PET scanner

MicroPET is a low-cost, high-resolution positron emission tomography (PET) scanner designed for imaging small animals. MicroPET operates exclusively w ithout septa, acquiring fully three-dimensional (3-D) data sets. The perfor mance of the projection-reprojection (3DRP), variable axial rebinning (VARB ), single slice rebinning (SSRB), and Fourier rebinning (FORE) methods for reconstruction of microPET data were evaluated. The algorithms were compare d with respect to resolution, noise variance, and reconstruction time. Resu lts suggested that the 3DRP algorithm gives the best combination of resolut ion and noise prformance in 9 min of reconstruction time on a Sun UltraSpar c I workstation. The FORE algorithm provided the most acceptable accelerate d method of reconstruction, giving similar resolution performance with a 10 %-20% degradation in noise variance in under 2 min. Significant degradation in the axial resolution was measured with the VARB and SSRB methods, offse tting the decrease in reconstruction time achieved with those methods. In-p lane angular mashing of the 3-D data before reconstruction led to a 50% red uction in reconstruction time but also introduced unacceptable tangential b lurring artifacts, This thorough evaluation of analytical 3-D reconstructio n techniques allowed for optimal selection of a reconstruction method for t he diverse range of microPET applications.