*2.2. CT Performance Phantom*

We used the CIRS Model 610 American Association of Physicists in Medicine (AAPM) CT performance phantom to measure spatial resolution and noise property. The CT number linearity insert (Part No. 610-02), which includes five cylinders with different densities, was a targeted imaging object for resolution measurement. The detailed specifications of the inserted cylinders are given in Table 3. Each cylinder has the same size and shape and has a low contrast against the background material, thus presenting a small absolute signal difference between the two materials. The larger the material index, the smaller the difference between the background and target material densities. Note that a small absolute difference between the densities of two materials does not always guarantee a small image contrast because the CT numbers are represented by the linear attenuation coefficients which are dependent on both X-ray energy and density.

**Table 3.** Material index and name of each cylinder embedded in the American Association of Physicists in Medicine (AAPM) phantom.


\* Poly methyl methacrylate (PMMA).

The insert (Part No. 610-01-05) is comprised of a uniform material with an aluminum pin at the center, and is a good candidate for measuring noise power. We assumed that the noise behaviors were the same for all materials because quantum and electronic noise, which are both stochastic events, are dominant over the entire area.
