A Novel Phantom for Evaluating Contrast Resolution in Single Photon Emission Computed Tomography (SPECT)

Abstract

We have designed a phantom to evaluate acquisition and reconstruction parameters using contrast transfer function (CTF). The goal of this study was to evaluate the accuracy of the phantom for contrast resolution. The phantom consisted of spaced (0-14 mm, 1 mm intervals) pairs of cubic containers (5 mm wide, 20 mm long, and 50 mm high). The phantom's accuracy was examined by comparing a real value of a measured count profile using the phantom with a theoretical value obtained by the line spread function (LSF) using a line source. A SPECT acquisition was 256×256 matrix (pixel size: 0.9×0.9 mm(2)), 360 degrees at 30 s/view. The radius of rotation was set to 15 cm, and the types of collimators were low energy high resolution (LEHR) and low middle energy general purpose (LMEGP). Reconstructions were performed with filtered backprojection and ordered subsets expectation maximization method (10 iterations, 10 subsets) with collimator-detector response correction. The actual measured count profile and CTF accorded closely with the theoretical one. The line pair (LP) phantom, obtained with smaller pixels, was really accurate. The size of cubic containers poses as a minimal problem for accurately evaluating the contrast resolution and plotted counts profile. This phantom could be a useful method for evaluating acquisition and reconstruction parameters in SPECT.