@misc{oai:repo.qst.go.jp:00070814,
 author = {Tashima, Hideaki and Yoshida, Eiji and Shinaji, Tetsuya and Hirano, Yoshiyuki and Kinouchi, Shoko and Nishikido, Fumihiko and Suga, Mikio and Haneishi, Hideaki and Ito, Hiroshi and Yamaya, Taiga and 田島 英朗 and 吉田 英治 and 品地 哲弥 and 平野 祥之 and 木内 尚子 and 錦戸 文彦 and 菅 幹生 and 羽石 秀昭 and 伊藤 浩 and 山谷 泰賀},
 month = {Jun},
 note = {Objectives: We are developing the OpenPET which can provide an accessible open space to the patient during PET scanning. In addition, we have proposed the real-time imaging system for the OpenPET toward PET-guided tumor-tracking radiation therapy. Although we demonstrated its tracking ability using a point source and a small OpenPET prototype, tumor tracking in human body is still challenging task because of its background activity when we use the 18F-FDG which is the best available tracer for tumors. In this study, we access conditions that the tumor tracking is feasible in human body by using the 4D XCAT phantom which is a realistic 4D human whole body phantom.
Methods: For generating 4D XCAT phantom with appropriate 18F-FDG distribution, SUVs were assigned to normal organs from literature. Spherical lesion with the SUV of 5 was placed at the right upper lung moving with respiratory cycle of 5 s. 4D Image was generated with the frame rate of 2 fps. Then, projection data of each frame were generated by forward projection with the system matrix including detector response functions of the human-sized OpenPET geometry having a gap of 300 mm. Detector blocks were 16x16x4 arrays of 3.0x3.0x7.5 mm3 scintillators. Attenuation of the body was included in the simulation, but scattering was ignored. Assuming the 18F-FDG injection of 370 MBq and time frame of 0.5 s and taking the system sensitivity into account, total event count in each projection data was adjusted to 500 k counts and then Poisson noise was added. 4D image was reconstructed frame by frame using the 3D OSEM method. Tumor position was extracted from the reconstructed 4D image by a pattern matching method.
Results: Tumor was moved 17 mm at the maximum with respiratory motion. Average error of the tumor position extracted from the reconstructed image was 2.7 mm, which was similar to the PET resolution.
Conclusions: We showed that tumor tracking by the OpenPET is feasible even in a realistic case.
Research Support: This work was partially supported by the Grant-in-Aid for Scientists Research (A) of Kakenhi (22240065) and by a research grant from the Association for Nuclear Technology in Medicine., SNM 2012 Annual Meeting},
 title = {Simulation study of real-time tumor tracking by the OpenPET using 4D XCAT phantom with 18F-FDG distribution},
 year = {2012}
}