Iterative reconstruction algorithm for half-cover dual helical cone-beam computed tomography

摘要

Cone-beam, computed tomography (CT) can be used for rapid volumetric imaging with high longitudinal resolution and is used extensively in a variety of non-destructive testing (NDT) and non-destructive evaluation (NDE) applications. But the field of view (FOV) of cone-beam CT is limited because of the limited size of the panel detector. In order to enlarge the FOV of cone-beam CT several scan modes are proposed, one of which is the half-cover dual helical scan mode. In half-cover dual helical cone-beam CT, the inspected object is scanned by two transversely truncated helical scans and two sets of truncated projections, which are complementary to each other, are acquired. Dual helical cone-beam CT can not only enlarge the FOV longitudinally, but also transversely. One author of this paper has proposed extended FDK and Katsevich-type algorithms for dual helical cone-beam CT. But both of the two algorithms are approximate algorithms. So, inevitably, there are slight artifacts (including drifting of the CT values) appearing in the reconstructed images, which may affect the subsequent interpretation of the reconstructed images. In order to improve the image quality further, we propose a modified SART iterative algorithm for dual helical cone-beam CT. Generally speaking, iterative reconstruction algorithms are applicable to any scan mode, however, direct SART reconstruction may introduce artifacts in the images because of data truncation in the projections. So, we make some modifications to the SART method to avoid artifacts in the images. Numerical experiments verify the proposed algorithm and the reconstruction results show that the images reconstructed by the modified SART algorithm are superior to those reconstructed by the extended FDK and Katsevich-type algorithms. Another advantage of the proposed algorithm is that it can make full use of the obtained projections rather than only parts of the projections.