Evaluation of Image Quality of the Heart Spect by Modifying the Imaging Protocol: A Phantom Study

Abstract

Introduction: Unfortunately, despite the low cost and non-invasive character of cardiac SPECT imaging, this method has
relatively low quality images. As a result, the sensitivity and specificity of this method is considerably lower than
angiography. In this study, which was performed using Monte Carlo simulation, the effects of increasing the number of
projection from 32 to 64 at the constant counting time on the image quality parameters were investigated.
Materials and methods: In this study, Tc99m was considered as accurate information of the ENSDF data bank. NCAT
humanoid digital phantom of a millimeter-resolution was used for simulation in three different sizes [large, medium and
small]. [GATE] code was used for SPECT images simulation. Two sets of images were taken from each phantom. In the
first shooting, 32 images with an average count of 100,000 as the witness images and in the second image, 64 images with
the count of 50,000 as the test images were taken. At the end, all images were reconstructed using FBP method and the
parameters of resolution, contrast and signal to noise ratio in the images were calculated to compare.
Results: By increasing the number of projection at the constant imaging time, sensible improvements in wall-to-cavity and
wall-to-background contrasts have been observed in the reconstructed. Since the resolution reflects contrast, improvement
of resolution is also expected. However, signal to noise ratio was decreased.