數值計算系統響應矩陣輔助雙平板正子攝影之影像重建與效能分析

正子攝影為一高靈敏度之核子醫學影像系統，其已被廣泛應用於臨床診斷如腫瘤偵測以及各項學術研究等。近年來針對藥物研究之動物實驗開發出一系列小動物專用之正子攝影影像系統，此類系統擁有高解析度之優勢，但由於所對應之待測體較小，其靈敏度難以提升至一般人體專用正子攝影的水準。雙平板式正子攝影為一裝配兩片大尺寸感測平板之小動物正子攝影系統，其除了保持高解析度的優勢外，所採用的大面積、高元件密度的平板式感測器矩陣更使其擁有高靈敏度的特性。然而雙平板式正子攝影的感測元件構造使其產生了嚴重的景深交互誤差，進而影響其解析度。本研究將採用由Robert L. Siddon提出之數值方法來計算雙平板式正子攝影之系統矩陣，雖然目前常用之蒙地卡羅演算法已能就雙平板式正子攝影系統模擬出一準確的系統響應矩陣，但此演算法在計算上相當費時以致於難以廣泛使用。本研究利用之演算法將以景深交互誤差作為考量依據，快速地計算出雙平板正子攝影之系統響應函數，進而得到系統響應矩陣。為了在計算上能有高效能的表現，在系統響應矩陣模擬以及重建演算法上將採用平行運算技術，以圖形處理器的多核心架構來處理大部分計算量，大幅度提升計算效能。我們將分別就模擬與實際的成相以本研究所計算之系統響應矩陣來進行重建，並與以蒙地卡羅演算法模擬之系統響應矩陣來重建之結果進行比較。此外也將就雙平板正子攝影系統對於指定信號來源的檢測能力以統計方法進行量化分析，藉此建立一針對個案進行影像系統最佳化之評斷流程。

關鍵字

雙平板式正子攝影；景深交互誤差；圖形處理器

並列摘要

Positron emission tomography (PET) is an imaging technique that has been widely employed in clinical diagnosis and research studies. Dual-head flat-panel PET is the dedicated PET imaging system designed specifically for small animals. Two large flat-panel detectors arranged in the compact scanning geometry can provide high detection sensitivity. However, the employment of narrow and thick detector crystals and large axial field-of-view can result in substantial parallax error, thus reducing the image resolution. In this work, we adopted this specific detector arrangement and computed the corresponding numerical system response functions that account for the parallax error. Although Monte-Carlo simulation can provide a system response matrix of high accuracy, it is too time-consuming to be employed in practice. Here we aimed to circumvent this implementation difficulty by applying Siddon’s algorithm to take into account of the blurring factors and then to generate the system response matrix. All simulations, including the system response matrix generation and image reconstruction, were accelerated by the massively parallel computing power of the graphics processing unit (GPU). The computational efficiency and the quality of reconstructed image were compared with that obtained by use of the Monte-Carlo simulated system response matrix. Moreover, the signal detection performance of the reconstructed images was also investigated by exploiting the statistical detection theory. Computer simulation studies were carried out to validate and quantitatively evaluate the proposed method.

並列關鍵字

Dual-head flat-panel PET ； parallax error ； Siddon’s algorithm ； graphics processing unit

參考文獻

Branco, S., S. Jan and P. Almeida. 2007. Monte Carlo simulations in small animal PET imaging. Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment. 580: 1127-1130.

Chou, C.-Y. and M. A. Anastasio. 2010. Noise texture and signal detectability in propagation-based x-ray phase-contrast tomography. Med. Phys. 37: 270-281.

Chou, C.-Y., Y.-Y. Chuo, Y. Hung and W. Wang. 2011. A fast forward projection using multithreads for multirays on GPUs in medical image reconstruction. Med. Phys. 38: 4052-4065.

Cutler, D. P., S. R. Cherry, E. J. Hoffman, W. M. Digby and M. E. Phelps. 1992. Design features and performance of a PET system for animal research. The Journal of Nuclear Medcine. 33: 595-604.

DeBenedetti, S., C. E. Cowan, W. R. Konneker and H. Primakoff. 1950. On the angular distribution of two-photon annihilation radiation. Physical Review. 77: 205.

國際替代計量

數值計算系統響應矩陣輔助雙平板正子攝影之影像重建與效能分析

全文下載

主題瀏覽