IEEE Trans Med Imaging. 2019 Dec;38(12):2735-2743. doi: 10.1109/TMI.2019.2912137. Epub 2019 Apr 22.
X-ray fluorescence CT (XFCT) has shown promise for molecular imaging of gold nanoparticles. To date, XFCT has been induced by kilovoltage photon beams due to the high photoelectric interaction probability. We compare K-shell and L-shell XFCT induced by photon, electron, and proton beams for two phantom sizes. A 2.5 and 5.0-cm diameter phantom with four 5 mm and 10 mm vials, respectively, with gold-solutions of 0.1%-2% by weight were built in TOPAS, a GEANT4-based Monte Carlo simulation tool. The 2.5-cm phantom was imaged with XFCT induced by beams of 7.45×10 81 keV- and 5 MeV-photons, 220 kVp- and 6 MV-photons, 10 MeV- and 100 MeV-electrons, and 100 MeV- and 250 MeV-protons. The doses between each phantom size were equal. First-generation CT geometry with 0.5 mm × 0.5 mm pencil beams with 0.5 mm-translation and 2°-rotation steps over each phantom was modeled. The scattered x-rays were detected on an idealized spherical detector from which the K-shell and L-shell fluorescent x-rays were extracted in 0.5 keV and 0.2 keV bins. XFCT images were generated using iterative reconstruction algorithms. The highest gold sensitivity was seen in the 81 keV-photon K-shell and L-shell images (0.004% and 0.007%) of the 5.0 cm-phantom at 30 mGy. For the 2.5 cm-phantom, the detection limits were 0.006%, 0.62%, and 0.28% for 81 keV-photon K-shell, 100 MeV-electron K-shell, and 100 MeV-proton L-shell images, respectively. The mean imaging dose was approximately 2-3 orders of magnitude higher in electron- and proton-XFCT compared to 81keV-photon XFCT. Our MC study demonstrates that the small-object XFCT imaging achieves the best performance when induced with kilovoltage-photon beams. Due to high imaging doses, electron- and proton-induced XFCT might be feasible for guiding nanoparticle-enhanced charged-particle radiotherapy.
X 射线荧光 CT(XFCT)在金纳米粒子的分子成像中显示出了前景。迄今为止,由于光电相互作用的概率很高,XFCT 是由千伏级光子束诱导产生的。我们比较了光子、电子和质子束诱导的 K 壳层和 L 壳层 XFCT,用于两种不同大小的模拟体模。在基于 GEANT4 的蒙特卡罗模拟工具 TOPAS 中构建了一个直径为 2.5 和 5.0 厘米的模拟体模,分别带有四个 5 毫米和 10 毫米的小瓶,内装金溶液的重量比为 0.1%-2%。使用 7.45×10 81 keV 和 5 MeV 光子、220 kVp 和 6 MV 光子、10 MeV 和 100 MeV 电子以及 100 MeV 和 250 MeV 质子束对 2.5 厘米模拟体模进行了 XFCT 成像。每个模拟体模的剂量相同。采用第一代 CT 几何形状,采用 0.5 毫米×0.5 毫米的铅笔束,在每个模拟体模上进行 0.5 毫米的平移和 2°的旋转步长。在理想化的球形探测器上检测散射 X 射线,并在 0.5 keV 和 0.2 keV -bin 中提取 K 壳层和 L 壳层荧光 X 射线。使用迭代重建算法生成 XFCT 图像。在 30 mGy 时,5.0 厘米模拟体模的 81 keV 光子 K 壳层和 L 壳层图像(0.004%和 0.007%)的金灵敏度最高。对于 2.5 厘米模拟体模,81keV 光子 K 壳层、100 MeV 电子 K 壳层和 100 MeV 质子 L 壳层图像的检测限分别为 0.006%、0.62%和 0.28%。与 81keV 光子 XFCT 相比,电子和质子 XFCT 的平均成像剂量大约高 2-3 个数量级。我们的 MC 研究表明,在千伏级光子束诱导下,小物体 XFCT 成像可获得最佳性能。由于成像剂量高,电子和质子诱导的 XFCT 可能可用于指导纳米粒子增强的带电粒子放射治疗。