This title appears in the Scientific Report :
2003
Please use the identifier:
http://hdl.handle.net/2128/1718 in citations.
Please use the identifier: http://dx.doi.org/10.1109/TNS.2003.817409 in citations.
Design optimization of the PMT-ClearPET prototypes based on simulation studies with GEANT 3
Design optimization of the PMT-ClearPET prototypes based on simulation studies with GEANT 3
Within the Crystal Clear Collaboration (CCC), four centers are developing second generation high performance small animal positron emission tomography (PET) scanners for different kinds of animals and medical applications. The first prototypes are photomultiplier tube (PMT)-based systems including d...
Saved in:
Personal Name(s): | Heinrichs, U. |
---|---|
Pietrzyk, U. / Ziemons, K. | |
Contributing Institute: |
Institut für Medizin; IME Zentralinstitut für Elektronik; ZEL |
Published in: | IEEE transactions on nuclear science, 50 (2003) S. 1428 - 1432 |
Imprint: |
New York, NY
IEEE
2003
|
Physical Description: |
1428 - 1432 |
DOI: |
10.1109/TNS.2003.817409 |
Document Type: |
Journal Article |
Research Program: |
Neurowissenschaften |
Series Title: |
IEEE Transactions on Nuclear Science
50 |
Subject (ZB): | |
Link: |
Get full text OpenAccess |
Publikationsportal JuSER |
Please use the identifier: http://dx.doi.org/10.1109/TNS.2003.817409 in citations.
Within the Crystal Clear Collaboration (CCC), four centers are developing second generation high performance small animal positron emission tomography (PET) scanners for different kinds of animals and medical applications. The first prototypes are photomultiplier tube (PMT)-based systems including depth of interaction (DOI) detection by using a phoswich layer of lutetium oxyorthosilicate (LSO) and lutetium yttrium aluminum perovskite (LuYAP). The aim of these simulation studies is to optimize sensitivity and spatial resolution of given designs, which vary in fields of view (FOVs) caused by different detector configurations (ring/octagon) and sizes. For this purpose the simulation tool GEANT3 (CERN, Geneva, Switzerland) was used.The simulation has shown that all PMT designs with no homogeneous axial coverage obviously have a very nonlinear corresponding axial sensitivity profile. By shifting every other detector module by a quarter of a PMT length in axial direction the sampling of the FOVs became more homogeneous. Applying an energy threshold of 350 keV, the regression coefficient increases from 0.818 for the nonshifted to 0.993 for the shifted design. Simulating a point source centered in the FOV (threshold: 350 keV) resulted in sensitivities of 4.2% for a 4 x 20PMT (LSO/LuYAP a 10 mm) and 3.8% for a 4 x 16PMT (LSO/LuYAP A 8 mm) ring design. The 3-D MLEM reconstruction of a point source shows the enormous improvement of resolution using a crystal double layer with DOI [3.1 mm at 40 mm from center FOV (CFOV)] instead of a 20 mm single layer (7.1 mm). |