DocumentCode
1802182
Title
A scatter correction based on the deconvolution method in SPECT
Author
Yang, Jiwei ; Karjalainen, Pasi ; Kuikka, Jyrki ; Patomäki, Lauri
Author_Institution
Dept. of Clinical Physiol. & Nucl. Med., Kuopio Univ. Hospital, Kuopio, Finland
Volume
1
fYear
1996
fDate
14-18 Oct 1996
Firstpage
501
Abstract
A scatter correction using a two-dimensional (2D) deconvolution method has been developed which compensates for Compton scattering in a triple-head gamma camera images. It assumes that the scatter distribution in any projection can be estimated from sinogram activity distribution using a position-invariant scatter kernel. The scatter distribution function has been derived from the measurements of a point source in a plastic phantom. The scatter correction technique is based on filtering by 2D convolution in the spatial domain. The spatial filter was designed to implement a 2D IIR filter. The quality of the compensation is also evaluated by comparing the ratio of the average counts/pixel with and without scatter correction in clinical brain blood flow studies
Keywords
Compton effect; IIR filters; blood flow measurement; brain; convolution; filtering theory; gamma-ray applications; medical image processing; single photon emission computed tomography; spatial filters; two-dimensional digital filters; 2D IIR filter; 2D deconvolution method; Compton scattering; SPECT; average counts/pixel ratio; clinical brain blood flow studies; measurements; plastic phantom; point source; position-invariant scatter kernel; projection; scatter correction; scatter distribution function; sinogram activity distribution; spatial domain filtering; spatial filter; triple-head gamma camera images; Cameras; Convolution; Deconvolution; Distribution functions; Filtering; IIR filters; Imaging phantoms; Kernel; Plastics; Scattering;
fLanguage
English
Publisher
ieee
Conference_Titel
Signal Processing, 1996., 3rd International Conference on
Conference_Location
Beijing
Print_ISBN
0-7803-2912-0
Type
conf
DOI
10.1109/ICSIGP.1996.567312
Filename
567312
Link To Document