Region of Interest Reconstruction from Dose-Minimized Super Short Scan Data

In this paper, we investigate an ROI image reconstruction in super short scan. The redundant rays in such scan are further shielded by using dynamic collimation. The acquired data coming with an extremely minimized radiation dose are sufficient for image reconstruction. For compensating resulting truncation, we apply a recently proposed algorithm Approximated Truncation Robust Algorithm for Computed Tomography (ATRACT). The evaluation with two clinical datasets demonstrates that high image quality is achieved using this super short scan, with more dose reduction compared to a standard short scan.