Title :
Local convolution in ectomography
Author :
Dale, Susanne M. ; Edholm, Paul R.
Author_Institution :
Dept. of Med. Imaging, Karolinska Inst., Huddinge, Sweden
Abstract :
In tomographic imaging, using limited angular sampling, details outside the imaged section are displaced along circles of blur. In ectomography, these are eliminated by a spatial convolution process. It is shown that the convolution function has to be as long as the projected dimension of the imaged object perpendicular to the section and twice the dimension parallel to the section.
Keywords :
patient diagnosis; blur circles; convolution function; ectomography; imaged object; limited angular sampling; local convolution; medical diagnostic imaging; projected dimension; spatial convolution process; tomographic imaging; Computer simulation; Convolution; Filters; Filtration; Frequency; Image reconstruction; Image sampling; Nuclear medicine; Shape; Tomography; Image Processing, Computer-Assisted; Mathematical Computing; Tomography, X-Ray Computed;
Journal_Title :
Biomedical Engineering, IEEE Transactions on
