Extension of the 4D NCAT phantom to dynamic X-ray CT simulation
The 4D NCAT phantom, developed to provide a realistic model of the human anatomy and physiology, is widely used in SPECT imaging research, but lacks anatomical details for application to high-resolution imaging such as x-ray CT and MRI. At the same time, current phantoms used in these areas lack the sufficient realism and flexibility to depict the complex shapes of real human organs and the deformations of those shapes due to anatomical variation and normal physiologic motion. We seek to fill that void by building upon the existing 4D NCAT phantom. Tissues distinguishable in x-ray CT imaging, including small details such as the pulmonary artery tree in the lungs and vasculature in the organs and body, were modeled based on segmented Visible Human CT data and high-resolution multi-slice spiral CT (MSCT) data. The models were created using a combination of NURBS and subdivision (SD) surfaces. The anatomy was extended to include structures in the head and abdomen for brain and prostate imaging applications. To efficiently simulate high-resolution x-ray CT images, we developed a new analytic projection algorithm to accurately calculate the projections (parallel, fan, or cone-beam geometries) directly from the surface definition of the phantom without using voxelization. The projection data were reconstructed using algorithms developed in our laboratory for fan- and cone-beam reconstruction of x-ray CT projection images. We conclude that the new 4D NCAT with its enhanced anatomy and physiology will provide an important tool in high-resolution imaging research.
