小動物實驗用針孔單光子發射斷層掃描之三維統計影像重建原理

Statistical 3D Image Reconstruction in Small Animal Pinhole SPECT

黃柏嘉(Po-Chia Lin) ； 許靖涵(Ching-Han Hsu)

《中華放射線技術學雜誌》 28卷1期 (2004/06) Pp. 29-39

https://doi.org/10.30045/CJRT.200406.0004

摘要

小動物實驗用針孔單光子發射斷層掃描，是利用針孔成像俱備影像放大的原理，與單光子發射斷層掃描結合。若搭配機率形式的統計模型，描述光子發射與接收的過程，統計影像重建法可重建出全三維、高解析度的小動物功能性照影影像，有助於提昇定性與定量分析的正確性。本文使用一最大相似度與均值最大化演算法，針對針孔單光子發射斷層掃描的硬體設計，與電腦模擬假體影像，進行全三維針孔單光子發射斷層掃描的影像重建分析。結果顯示使用統計為基礎的疊代式影像重建演算法，可提供重建高解析度與高正確性的影像。對於需要高品質影像的小型動物實驗而言，驗證疊代式影像重建法應用在針孔單光發射層掃瞄的可行性與正確性。

關鍵字

針孔單光子發射斷層掃描 ； 最大相似度與均值最大化演算法 ； 小動物影像

並列摘要

Functional imaging of small animal studies plays an important role in the development of new diagnostic and therapeutic pharmaceuticals. Small animal pinhole SPECT is a powerful imaging modality which integrates the image amplification property of pinhole with SPECT scanner. With an accurate statistical modeling describing the photon detection process, statistical image reconstruction can produce high-resolution 3D SPECT image results for various quantitative and qualitative analyses. In this work, we describe a statistical framework for fully 3D image reconstruction based on a general pinhole SPECT configuration of small animal. Using computer-simulated phantom studies, the experimental results indicate that the proposed framework can produce high-resolution image results. It is anticipated the proposed image reconstruction framework should be applicable to either clinical pinhole SPECT systems, or some dedicated pinhole SPECT scanners for small animal imaging.

並列關鍵字

Pinhole SPECT ； Maximum Likelihood Expectation Maximization ； Small Animal Imaging