Noise properties of low-dose CT projections and noise treatment by scale transformations

Projection data acquired for image reconstruction of low-dose computed tomography (CT) are degraded by many factors. These factors complicate noise analysis on the projection data and render a very challenging task for noise reduction. In this study, we first investigate the noise property of the projection data by analyzing a repeatedly acquired experimental phantom data set, in which the phantom was scanned 900 times at a fixed projection angle. The statistical analysis shows that the noise can be regarded as normally distributed with a nonlinear signal-dependent variance. Based on this observation, we then utilize scale transformations to modulate the projection data so that the data variance can be stabilized to be signal independent. By analyzing the relationship between the data standard deviation and the data mean level, we propose a segmented logarithmic transforms for the stabilization of the non-stationary noise. After the scale transformations, the noise variance becomes approximately a constant. A two-dimensional Wiener filter is then designed for an analytical treatment of the noise. Experimental results show that the proposed method has a better noise reduction performance without circular artifacts, by visual judgment, as compared to conventional filters, such as the Hanning filter.