Title :
Local threshold for segmented attenuation correction of PET imaging of the thorax
Author :
Xu, M. ; Luk, W.K. ; Cutler, P.D. ; Digby, W.M.
Author_Institution :
Mallinckrodt Inst. of Radiol., St. Louis, MO, USA
fDate :
31 Oct-6 Nov 1993
Abstract :
Local threshold for segmented attenuation correction technique has been developed for positron emission tomography using short (2-3 minutes) post-injection transmission scans. The technique implements an optimal threshold method on localized histograms to get pseudo-anatomic segmentation on transmission images. Theoretical values of attenuation coefficient are assigned to correspondent anatomic regions. Emission images are reconstructed using attenuation correction factors computed by forward-projecting segmented transmission images. Phantoms and clinical cardiac images are studied using this technique. The technique has been found linearly correlated with the standard method (15 minutes, pre-injection transmission scans), reducing noise in the final emission Images, eliminating process for substraction emission data from transmission scan, cutting down scan time and increasing patient throughput
Keywords :
gamma-ray absorption; image segmentation; medical image processing; positron emission tomography; 15 min; 2 to 3 min; PET imaging; attenuation coefficient; clinical cardiac images; correspondent anatomic regions; forward-projecting segmented transmission images; local threshold; medical diagnostic imaging; noise reduction; nuclear medicine; patient throughput increase; phantoms; preinjection transmission scans; scan time reduction; segmented attenuation correction; substraction emission data; thorax imaging; Attenuation; Biological tissues; Histograms; Image segmentation; Lungs; Pollution measurement; Positron emission tomography; Radiology; Thorax; Throughput;
Conference_Titel :
Nuclear Science Symposium and Medical Imaging Conference, 1993., 1993 IEEE Conference Record.
Conference_Location :
San Francisco, CA
Print_ISBN :
0-7803-1487-5
DOI :
10.1109/NSSMIC.1993.373615