The long object imaging problem in volumetric CT using iterative reconstruction algorithm, revisited

The long object imaging problem in volumetric CT using iterative reconstruction algorithm is a challenge due to the intrinsic longitudinal (z-axis) truncation in the circular scanning trajectory. We revisited the problem in this work and proposed a simplified version of our original approach. We simply removed the projection data expansion and empirical weighting scheme in our original method while reconstructed only the image in an extended region. It seems that error will be introduced in the reconstructed image volume from those rays traveling both through the region of interest (ROI) and outside ROI. Hence error will propagate further into the image volume resulting the useful ROI diminishes for each iteration. However, our results showed that the image quality was well preserved by use of iterative reconstruction method. Artifacts due to the longitudinal truncation did not propagate as originally expected. We demonstrated our results of two sets of clinical data, head and abdomen, respectively, acquired from Aquilion ONE™ CT scanner (Toshiba American Medical Systems, Tustin, CA, USA) reconstructed by use of iterative reconstruction algorithm.