|
KMID : 1147120170230010007
|
|
Journal of the Korean Society of Imaging Informatics in Medicine
2017 Volume.23 No. 1 p.7 ~ p.13
|
|
|
4D SPECT Image Reconstruction using MAP Estimation
|
|
Cho Sang-Hoon
So Jeong-Tae
Chang Ji-Eun
Kim Chul-Han
Cho Seung-Ryong
|
|
|
Abstract
|
|
|
Background: SPECT is widely used to see the biochemical processes by uptake of radiopharmaceuticals or tracers, of which 99mTc is commonly used due to its adequate half-life, about 6.03 hours. However, compared to PET, dynamic studies have not been active in SPECT because of several factors such as weak signals and slow scanning time with the use of collimator. The purpose of our study was to reconstruct 4D SPECT images with reduced elongation artifacts caused by object motion.
Materials and Methods: 60 projection data of linearly moving respiratory chest-phantom, where radio-isotope (99mTc) was inserted, were acquired by xSPECT with two opposing head detectors. After computing center of mass for each projection data, projections were separated into each phase by spatial phase sorting. For image reconstruction, separated data were fed into the Maximum a posteriori (MAP) estimation algorithm regularized by Gibbs prior.
Results: Target motions were modeled with the center of mass and variations of the value were reduced by using nonlinear least square. 4D SPECT image was reconstructed while reducing elongation artifact which arise from fast moving imaged target by using blur matrix.
Conclusion: The proposed method reduced elongation artifacts of a moving target and reconstructed 4D SPECT image using MAP estimation.
|
|
|
KEYWORD
|
|
|
Maximum A Posteriori, Single photon emission tomography (SPECT), Time resolving, Nuclear medicine, 4D SPECT, Elongation artifact
|
|
|
FullTexts / Linksout information
|
|
|
|
|
|
Listed journal information
|
|
|
|