@article{oai:gchs.repo.nii.ac.jp:00000037,
 author = {高橋, 康幸 and 織内, 昇 and 成田, みち子 and 勘崎, 貴雄 and 齋藤, 享子 and 北尾, 総一 and 遠藤, 啓吾 and 村瀬, 研也 and Takahashi, Yasuyuki and Oriuchi, Noboru and Narita, Michiko and Kanzaki, Takao and Saito, Kyoko and Kitao, Soichi and Endo, Keigo and Murase, Kenya},
 journal = {群馬県立県民健康科学大学紀要},
 month = {Mar},
 note = {Article, 目的:^<18>F-FDG PETにおけるRAMLA画像再構成法の緩和係数について検討した.方法:ファントム実験では^<18>Fを均一に満たした円柱ファントムを用い,内部には直径8,13,16,21,26,31mmサイズの陽性球を等間隔に設定した.陽性球の集積比は,体重60kgの患者で185MBq投与後60分を想定し,SUVが異常判定の基準とされる2.5(9.1kBq/ml)および高集積の5.0(18.2kBq/ml)を作成した.バックグラウンドは3.64kBq/mlである.なお,PET/CT装置はGemini GXL(PHILIPS)で,撮像は日常検査条件の1ペットポジションあたり2分と理想的条件と考えた20分を収集した.画像再構成は,Iterationを2回とするRAMLA法で,FWHMは5mmのガウス関数で平滑化した.この条件で緩和係数を0.02から0.5まで変化させ,画質は視覚評価とCVにより比較検討した.結果:SUV5.0における収集時間の違いでは,緩和係数が0.08より増加すると日常検査条件で陽性球の描出性が低下した.これは,SUV2.5においても同様の傾向がみられ,21mm以下の集積は認識できなかった.結論:適切な緩和係数は,収集条件によって異なることが示唆された., Objectives: The row-action maximum likelihood algorithm (RAMLA) method is a better image reconstruction method that can suppress the noise present in a PET image. This algorithm includes a relaxation parameter, and this study considered its validation in ^<18>F-FDG. Methods: A PET scanner Gemini GXL (PHILIPS) system was used in this study. In the phantom study, a cylindrical phantom with 6 hot spheres was filled with ^<18>F and arranged at equal intervals. The diameters of the hot spheres were 8, 13, 16, 21, 26, and 31 mm, respectively. The accumulation of this phantom was calculated to give standardized uptakes value (SUV) of 2.5 (9.1 kBq/ml) and 5.0 (18.2 kBq/ml) on images acquired 60 minutes after injection of 185 MBq into a patient weighing 60 kg. The radioactive concentration of the background is 3.64 kBq/ml to get an SUV of 1.0. The acquisition time used was for a routine acquisition of 2 minutes/bed position and an ideal condition of 20 minutes/bed position. The relaxation parameter was changed from 0.02 to 0.5. Image analyses included visual evaluation and statistical analysis. Results: On visual evaluation, image quality decreased when the relaxation parameter was larger than 0.08. SUV was inaccurate with 2 minutes acquisition. Image quality depends on the SUV difference. When the relaxation parameter was larger than 0.08, image quality degraded significantly. With an SUV of 2.5, accumulation was not recognized when the relaxation parameter was 0.5. Evaluation of the quality of image using the coefficient of variation (CV) showed that it deteriorated as the relaxation parameter increased. Conclusions: The results of the present study indicate that the optimal relaxation parameter differs according to the acquisition conditions., 原著, 群馬県立県民健康科学大学紀要 4,25-30,2009},
 pages = {25--30},
 title = {^<18>F-FDG PET における RAMLA画像再構成法の基礎的検討},
 volume = {4},
 year = {2009}
}