背景：微型正子放射電腦斷層造影(microPET)?藥物研發與塞因工程研究的一大利器。針對小動物的藥物動力學研究，microPET量化的準確性是其主要的挑戰。因此，一種有效的microPET散射矯正法更顯得重要。本研究之目的在於提出一套散射矯正演算法，應用於microPET造影上，以期達到率確的定量分析。方法：本研究在microPET上架設射束阻檔裝置(beam stopper device; BS)，並將小動物掃描程序分?放置射束檔塊器與無射束檔塊器兩部份。在假設散射分佈?一緩慢變化函數的前提下，根據兩次婦描與空氣婦描的資訊，可以求得被檔塊阻檔部份的散射分率，並利用cubic spline內插法回復出整個散射分佈，以達到散射矯正之目的。本研究利用核研所的microPET進行假體實驗，以SimSET蒙地卡羅模擬結果做?標準，比較BS方法與臨床上常用的單散射光子模擬法(single scatter simulation; SSS)的矯正成效。結果：BS方法評估的散射分率?15.24%，較接近蒙地卡羅的模擬結果。此外，BS法比SSS法更能有效預到散射分佈，並率確復原影像的對比值。結論：BS方法應用即時量測以估計散射分佈，可以考慮到照野外以及非物體本身的散射，因此更能率確評估散射量。

Background: The microPET system provides investigators a helpful tool for the drug development and the imaging of gene expression. Especially for pharmacokinetic studies on the small animal, the most challenging issue is the accurate quantification of images. Therefore, an effective and reliable scatter correction method is necessary to improve quantitative performance. We proposed a scatter correction method for the microPET system with the aim of estimating the scatter distribution exactly. Methods: We extend the use of the beam stopper device and divide the animal scan into two sub-scans with and without the beam stoppers. By assuming that the scattered radiation tends to be a spatially slow-varying function, the scatter components at LOR blocked by each stopper can be estimated and the whole scattered sinogram can be recovered using interpolation from these local measurements. The phantom study on the microPET was investigated and we also simulated the Monte Carlo result as a golden standard to compare with the performance of Beam Stopper (BS) method and Single Scatter Simulation (SSS) method used in clinic currently. Results: The scatter fraction evaluated by BS method was 15.24% and was close to the Monde Carlo result. Besides, the proposed method outperformed the SSS method in term of the estimated scatter distributions and the recovery of contrast. Conclusion: The proposed method considers scatter from outside of field of view and multiple scatter. Thus, it could improve the quantitative performance with microPET applications.