TY - GEN
T1 - PET resolution and image quality optimization study for different detector block geometries and DOI designs
AU - Zhang, Yuxuan
AU - Wong, Wai Hoi
AU - Liu, Shitao
AU - Liu, Jiguo
AU - Kim, Soonseok
AU - Baghaei, Hossain
AU - Li, Hongdi
AU - Ramirez, Rocio
AU - Wang, Yu
PY - 2007
Y1 - 2007
N2 - Detector geometry for PET camera is one of the most important factors that will determine the resolution and image quality of the camera. In this work, the point source resolutions and hot-rod phantom images of the PET systems with different detector geometries are studied using Monte Carlo simulations. The PET systems been studied include a human brain PET and an animal PET with the typical detector ring dimensions but different crystal sizes, number of layers and block geometries. The detector ring diameter is 480 mm for the brain PET and 160 mm for the animal PET. The blocks been studied include 1-layer and 2-layer geometries. Two types of 2layer blocks are studied, one has a half-crystal-offset (HCO) between the top and bottom layers to double the radial and axial samplings, and the other one is the regular (non-HCO) 2-layer block. The blocks in the brain PET is 40×40×20 mm3 and with a crystal matrix of 10×10, 13×13 and 15×15; the block in the animal PET is 20×20×10 mm3 and with a crystal matrix of 10×10, 12×12 and 14×14. Point source resolution curves (radial, tangential and axial) as the function of off-center distance are obtained with an F-18 source. Two Derenzo-like hot-rod phantoms are used for overall image quality test. The results show the significant improvement on the DOI effect with 2-layer block especially in the radial direction. Using the HCO block, the transaxial resolution is a little better than that of non-HCO block in the center region of the FOV but become worse with some distance from the center because of the sampling rates in different radial offset positions are not the same for these two types of blocks. Another important finding is that the axial resolution with HCO block is much better than that from 1-layer or 2-layer non-HCO block. Therefore using HCO block with larger crystal pitch can still achieve better image quality not only in transaxial plane at the outside region of a large FOV compare to the small crystal pitch with 1-layer block, but also better resolution in axial direction without additional cost, which might be a better choice for high-resolution PET camera design.
AB - Detector geometry for PET camera is one of the most important factors that will determine the resolution and image quality of the camera. In this work, the point source resolutions and hot-rod phantom images of the PET systems with different detector geometries are studied using Monte Carlo simulations. The PET systems been studied include a human brain PET and an animal PET with the typical detector ring dimensions but different crystal sizes, number of layers and block geometries. The detector ring diameter is 480 mm for the brain PET and 160 mm for the animal PET. The blocks been studied include 1-layer and 2-layer geometries. Two types of 2layer blocks are studied, one has a half-crystal-offset (HCO) between the top and bottom layers to double the radial and axial samplings, and the other one is the regular (non-HCO) 2-layer block. The blocks in the brain PET is 40×40×20 mm3 and with a crystal matrix of 10×10, 13×13 and 15×15; the block in the animal PET is 20×20×10 mm3 and with a crystal matrix of 10×10, 12×12 and 14×14. Point source resolution curves (radial, tangential and axial) as the function of off-center distance are obtained with an F-18 source. Two Derenzo-like hot-rod phantoms are used for overall image quality test. The results show the significant improvement on the DOI effect with 2-layer block especially in the radial direction. Using the HCO block, the transaxial resolution is a little better than that of non-HCO block in the center region of the FOV but become worse with some distance from the center because of the sampling rates in different radial offset positions are not the same for these two types of blocks. Another important finding is that the axial resolution with HCO block is much better than that from 1-layer or 2-layer non-HCO block. Therefore using HCO block with larger crystal pitch can still achieve better image quality not only in transaxial plane at the outside region of a large FOV compare to the small crystal pitch with 1-layer block, but also better resolution in axial direction without additional cost, which might be a better choice for high-resolution PET camera design.
UR - http://www.scopus.com/inward/record.url?scp=48149102881&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=48149102881&partnerID=8YFLogxK
U2 - 10.1109/NSSMIC.2007.4436737
DO - 10.1109/NSSMIC.2007.4436737
M3 - Conference contribution
AN - SCOPUS:48149102881
SN - 1424409233
SN - 9781424409235
T3 - IEEE Nuclear Science Symposium Conference Record
SP - 2885
EP - 2888
BT - 2007 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS-MIC
T2 - 2007 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS-MIC
Y2 - 27 October 2007 through 3 November 2007
ER -