Comparison of two projectors with image-based resolution models in a fully 3D list-mode PET reconstruction

In this paper, an image-based and spatially variant resolution model (i.e. PSF model) was derived from reconstructions of a series of point sources, and used to recover the resolution loss in PET imaging. Since this PSF model was not only system dependent, but also reconstruction algorithm dependent, two reconstruction projectors were compared: one was the single-ray-tracing (1-ray) and another was the Area-Simulating-Volume (ASV). An IEC phantom and a female patient were reconstructed using a fully 3D time-of-flight list-mode ordered-subsets expectation-maximization algorithm with different projector-PSF combinations: 1-ray, 1-ray-PSF, ASV and ASV-PSF. Physical corrections such as attenuation, scatter, random, detector normalization, decay and dead time were also included in these reconstructions. For both phantom and patient datasets, projectors with resolution models demonstrate superior quantitation performance compared to those without resolution models in terms of resolution, contrast and noise. Results also show 1-ray-PSF has very comparable image qualities to ASV-PSF, though the latter calculates more accurate geometric probabilities in the system matrix. Therefore, a simple projector with an image-based PSF model might be suggested for the future reconstruction development, especially in reconstruction with a GPU acceleration.