National Institutes for Quantum Science and Technology, 4-9-1, Anagawa, Inage-ku, Chiba-shi, Chiba, 263-8555, Japan.
Medical Engineering Course, Graduate School of Science and Engineering, Chiba University, 1-33, Yayoicho, Inage-ku, Chiba-shi, Chiba, 263-8522, Japan.
Radiol Phys Technol. 2023 Jun;16(2):254-261. doi: 10.1007/s12194-023-00714-5. Epub 2023 Mar 21.
In Compton PET, that has a scatterer inserted inside a PET ring, there are multi-interaction events that can be treated as both PET and Compton events. A PET event from multi-interaction events that include a Compton event and a photoelectric absorption event or two Compton events can be extracted by applying a PET recovery method. In this study, we aimed to establish a method to maximize image quality by utilizing such redundant events. We conducted brain-scale Monte Carlo simulations of a C-shaped Compton-PET geometry and a whole gamma imaging (WGI) geometry. Images were reconstructed by a hybrid image reconstruction method combining both PET and Compton events. The result showed that the spatial resolution was improved when treated as PET events while keeping the noise level. The effect of improvement was more significant in WGI than in C-shaped Compton PET because the number of events recovered as PET events having more accurate spatial information was much larger in WGI. When the PET-recovered multi-interaction events were also included as Compton events in the hybrid reconstruction, we did not observe any improvement in image quality, while the number of used events was largest. The results suggested that treating events as PET events exclusively was better for image quality.
在 Compton PET 中，在 PET 环内插入了散射体，其中存在可以视为 PET 和康普顿事件的多相互作用事件。通过应用 PET 恢复方法，可以从包含康普顿事件和光电吸收事件或两个康普顿事件的多相互作用事件中提取 PET 事件。在这项研究中，我们旨在建立一种利用这种冗余事件来最大化图像质量的方法。我们对 C 形康普顿 PET 几何形状和全伽马成像 (WGI) 几何形状进行了大脑规模的蒙特卡罗模拟。通过结合 PET 和康普顿事件的混合图像重建方法对图像进行了重建。结果表明，在保持噪声水平的同时，将其视为 PET 事件可提高空间分辨率。在 WGI 中改善效果更为明显，因为在 WGI 中可以恢复出更多具有更准确空间信息的 PET 事件，从而可以恢复出更多的事件。当将 PET 恢复的多相互作用事件也作为混合重建中的康普顿事件包含在内时，我们没有观察到图像质量有任何改善，而使用的事件数量却是最多的。结果表明，仅将事件视为 PET 事件对图像质量更有利。
