@misc{oai:repo.qst.go.jp:00070601,
 author = {Shiraishi, Takahiro and Yoshikawa, Kyosan and Fukumura, Akifumi and Mizuno, Hideyuki and Tanimoto, Katsuyuki and Ishii, Noriyuki and Inagaki, Eri and Omatsu, Mika and Ohashi, Seiya and Toubaru, Sachiko and Kamada, Tadashi and Ando, Yutaka and Tsuji, Hiroshi and Watanabe, Kazuhiro and et.al and 白石 貴博 and 吉川 京燦 and 福村 明史 and 水野 秀之 and 谷本 克之 and 石井 徳幸 and 稲垣 枝里 and 尾松 美香 and 大橋 靖也 and 桃原 幸子 and 鎌田 正 and 安藤 裕 and 辻 比呂志 and 渡邊 和洋},
 month = {Oct},
 note = {[Purpose] Detecting the distribution of positron emitter generated by nuclear fragmentation reaction in carbon ion beam therapy (auto-activation) applied at Heavy Ion Medical Accelerator in Chiba (HIMAC) by means of positron emission tomography (PET) is novel method for therapy quality assurance especially assessing irradiated area. Since a new 3D irradiation system with pencil-beam scanning (fast- scanning system) has been developed at NIRS, and we are planning to assess the accuracy of the delivered dose distribution by clinical auto-activation PET study. In preparation for this purpose, we performed some phantom experiments for determine optimal scan duration and compare PET image quality of auto-activation PET study between conventional irradiation system and fast scanning system. 
\n[Materials and Methods] A SET-3000GCT/X (PET/CT; Shimadzu Corp.) dedicated to the 3D-acquisition mode is used for auto-activation PET imaging. It uses gadolinium oxyorthosilicate (GSO) detectors and provides 99 sections with a center-to-center distance of 2.6mm. To determine the optical scan duration, the phantom irradiated with carbon-ion beam is scanned for 60min in list-mode, then a 5min transmission scan with a 137Cs point source is performed. Before reconstruction, list-mode events are histogrammed as PET raw data obtained at each acquisition time (10, 20, 30, 40, 50 and 60min). The emission images are reconstructed with the manufacturer software using Fourier rebinning followed by 2D-dynamic-RAMLA. Reconstructed images are evaluated visually and normal mean square error (NMSE) is assessed. In addition, image noise is derived from region-of-interest (ROI) analysis. Image noise was defined as the coefficient of variation (COV, 100 x SD/mean [%]) of the pixel values within a homogeneous ROI. 
\n[Results] There is no obvious difference of image quality between 40min acquisition and 60min acquisition by visually evaluation. Based on statistical analysis, NMSE value and COV are 0.0017 and 12% at 40min acquisition, respectively, while NMSE value and COV are 0.0023 and 14% at 30 min acquisition, respectively. According to these results, we conclude the 40min acquisition as optimal scan duration for clinical auto-activation PET study. The distribution of positron emitters generated by carbon ion beam for a clinical target volume (CTV) is slightly larger in fast-scanning system compared with conventional irradiation system. The difference is all explained by presence of penumbra in fast-scanning system. These results show that auto-activation PET method can visualize precisely the absorbed radiation area. 
\n[Conclusions] According to our analysis, it is clarified that auto-activation PET method can detect precisely absorbed radiation area and it plays an important role in quality assurance of heavy ion beam therapy., EANM Annual Congress 2011},
 title = {Experimental PET/CT study of imaging the small amount of positron emitters generated by carbon ion beam irradiation},
 year = {2011}
}