AbstractID: 9760 Title: Image reconstruction with improved noise properties for asymmetric conebeam CT
Cone-beam CT with a displaced 2D detector array, which is referred to as
asymmetric cone-beam CT, can be used for increasing the size of the field of
measurement (FOM) in micro-CT or in megavoltage CT for patient localization in
radiation therapy. From the data acquired with such configuration, images can be
reconstructed by use of the conventional FDK algorithm with data weighted by a
smooth weighting function. With the displacement of detector and increase of the
FOM size, however, the FDK algorithm may significantly amplify data noise and
aliasing artifacts because it involves a spatially-variant weighting factor. In
this work, we propose an algorithm to reconstruct images from asymmetric conebeam data. This algorithm eliminates the spatially-variant weighting factor
involved in the FDK algorithm, and thus reconstructs images with improved noise
properties. It maintains the image spatial resolution by use of a shift-variant
filtration scheme to avoid interpolation along the radial direction.
Additionally, the proposed algorithm can compensate for the intensity-drop
effect resulted from the missing-data problem in the FDK algorithm. This
algorithm is particularly robust when applied to asymmetric cone-beam CT with
large size of FOM and/or small focal lengths.
Supported by the National Institutes of Health