Investigation of sparse-angle view in cone beam computed tomography (CBCT) reconstruction algorithm using a sinogram interpolaton method

Abstract

Several methods for reducing patient dose have been widely studied in computed tomography (CT). The radiation exposure dose for patient can be decreased with sparseangle view reconstruction in CT. In sparse-angle view reconstruction, limited projection images are taken around the object and these are incomplete projection data for obtaining the complete sinogram data. A sinogram interpolation method is the one of solutions to reconstruct the image in sparse-angle view. In this study, we applied a linear sinogram interpolation method in cone-beam computed tomography (CBCT) reconstruction. We simulated the CBCT system with the MATLAB R2012a program to obtain projection data and reconstruct images using three-dimensional version of the Shepp-Logan phantom. Root mean square error (RMSE) was measured in a sinogram interpolation method and normal Feldkamp-Davis- Kress (FDK) reconstruction. The lower RMSE factor means the reconstruction image was similar to the original image. In conclusion, our results demonstrated that the sinogram interpolation method can minimize the error with original image compared with normal FDK reconstruction.