Image reconstruction algorithms for half-scan cone beam computed tomography

Abstract

Modern CT and micro-CT scanners are rapidly moving from fan-beam towards cone-beam geometry. One of the important advantages of the cone-beam CT is its fast volumetric scanning capability. It provides the opportunity for in-vivo dynamic study, which is becoming more and more important. However, in wide cone angles, current existing approximate algorithms suffer from severe image artifacts due to the nature of their approximation and/or missing information. Therefore, the algorithms that have high temporal resolution and are free from image artifacts are highly demanded these days. In this study, we have developed a series of half-scan algorithms based on Grangeat`s formula as one of such efforts. Prior to the algorithmic development, we have explicitly revealed artifacts associated with implementation of Grangeat`s formula, which is the main framework of this work as well as many other exact reconstruction algorithms developed during the last decade. Although the exact cone-beam approach is theoretically error-free, it is subject to image artifacts due to the discrete nature of numerical implementation. We report a study on image artifacts associated with the Grangeat algorithm as applied to a circular scanning locus. Three types of artifacts are found, which are thorn, wrinkle, and V-shape artifacts. Half-scan CT algorithms are advantageous in terms of temporal resolution, and widely used in fan-beam and cone-beam geometry. While existing half-scan algorithms for cone-beam CT are in the Feldkamp framework, in this dissertation we compensate missing data explicitly in the Grangeat framework, and formulate half-scan algorithms in the circular and helical scanning for the untruncated short object case first. The half-scan spans 180 degree plus two cone angles that guarantee sufficient data for reconstruction of the mid-plane defined by the source trajectory. The smooth half-scan weighting functions are designed for suppression of data inconsistence. Numerical s...