If there is a missing part in projection data, the tomogram reconstructed by a CT algorithm drastically deteriorates. To alleviate this problem, a new method is proposed for reconstructing CT images from projection data with some parts missing. In this algorithm actual projection data and two types of known contour information (i. e. , the cross-sectional boundary and the outline of opaque parts of an object) are used. During each iteration of reconstruction and projection corrections on an estimate of missing data are made. When the difference between estimated projection data and measured projection data becomes smaller than a predetermined value, the reconstruction is considered to have converged.
|Number of pages||9|
|Journal||Systems, computers, controls|
|Publication status||Published - 1981|
ASJC Scopus subject areas