The Quantitative Effect of Noise and Object Diameter on Low-Contrast Detectability of AAPM CT Performance Phantom Images

Abstract

Parameters for determining computed tomography (CT) image quality include noise and low-contrast detectability. Studies on low-contrast detectability using the AAPM CT performance phantom have several limitations, such as the absence of quantitative information on the effect of noise and object size on low-contrast detectability. In this study, the quantitative effect of noise and object diameter on low-contrast detectability were investigated. Images of the American Association of Physicists in Medicine (AAPM) CT performance phantom model 610 were acquired with a tube voltage of 120 kV and tube currents of 50, 100, 150, and 200 mA. The low-contrast section of the AAPM CT performance phantom model 610 has objects with diameters between 2.5 and 7.5 mm. We analysed the mean CT number, noise level, signal-to noise ratio (SNR), and contrast-to-noise ratio (CNR), acquired using MatLab software. The results obtained indicate that noise and object size affect low-contrast detectability. The CNRs increase linearly with increasing of object diameter with R2 of 0.88, 0.67, 0.75, and 0.83 for tube currents of 50, 100, 150 and 200 mA, respectively.