Experimental comparison of lesion detectability for four fully-3D PET reconstruction schemes

The objective of this work was to evaluate the lesion detectability performance of four fully-3D PET reconstruction schemes using experimentally acquired data. A multicompartment anthropomorphic phantom was setup to mimic whole-body FDG cancer imaging and scanned twelve times in 3D mode. Eight of the scans had twenty-six ⁶⁸Ge "shell-less" lesions (6, 8, 10, 12, 16 mm diam.) placed throughout the phantom with various target:background ratios. This provided lesion-present and lesion-absent datasets with known truth appropriate for evaluating lesion detectability by localization receiver operating characteristics (LROC) methods. Four reconstruction schemes were studied: (1) Fourier rebinning (FORE) followed by 2D attenuation-weighted ordered-subsets expectation-maximization; (2) fully-3D AW-OSEM; (3) fully-3D ordinary-Poisson line-of- response (LOR-)OSEM; and (4) fully-3D LOR-OSEM with an accurate point-spread function (PSF) model. Two forms of LROC analysis were performed. First, a channelized non-prewhitened (CNPW) observer was used to optimize processing parameters (number of iterations, post-reconstruction filter) for the human observer study. Human observers then rated each image and selected the most-likely lesion location. The area under the LROC curve (A_LROC) and the probability of correct localization were used as figures-of-merit. The results of the human observer study found no statistically significant difference between FORE and AW-OSEM3D (A_LROC=0.41 and 0.36, respectively), an increase in lesion detection performance for LOR-OSEM3D (A_LROC=0.45, p=0.076), and additional improvement with the use of the PSF model (A_LROC=0.55, p=0.024). The numerical CNPW observer provided the same rankings in performance among algorithms, but obtained different values of A_LROC. These results show improved lesion detection performance for the reconstruction algorithms with more sophisticated statistical and ima- ging models as compared to the previous-generation algorithms.