A back-projection-and-filtering-like (BPF-like) reconstruction method with the deep learning filtration from listmode data in TOF-PET.
A back-projection-and-filtering-like (BPF-like) reconstruction method with the deep learning filtration from listmode data in TOF-PET.
复制标题
DOI：
10.1002/mp.15520
复制
发表时间：
2022-04
期刊：
影响因子：
3.8
通讯作者：
Huang, Qiu
中科院分区：
文献类型：
--
作者：
Lv, Li;Zeng, Gengsheng L.;Zan, Yunlong;Hong, Xiang;Guo, Minghao;Chen, Gaoyu;Tao, Weijie;Ding, Wenxiang;Huang, Qiu
飞行时间(TOF)信息提高了正电子发射断层扫描(PET)成像的信噪比(SNR)。现有的TOF-PET分析算法通常遵循从正弦图数据重建图像的滤波反投影过程。这项工作的目的是开发一种类似反投影和滤波(BPF-like)的算法，直接从列表模式数据快速重建TOF PET图像。我们将传统的二维非TOF PET投影模型扩展到TOF投影模型，其中投影数据表示为沿投影方向由一维TOF核加权的线积分。在推导列表模式数据的中心切片定理和TOF反投影的基础上，设计了一种改进的U-Net结构的深度学习网络来执行空间过滤(重构过滤器)。用三种不同时间分辨率的TOF PET列表模式数据对两种活动模体进行了蒙特卡罗模拟，验证了BP网络方法的有效性。该网络只在模拟的全剂量xCAT数据集上进行了训练，然后在不同时间分辨率和剂量水平的xCAT和Jaszczak数据上进行了评估。重建图像表明，与传统的BPF算法和提出的TOF PET的MLEM算法相比，BP网络方法在峰值信噪比、相对均方误差和结构相似性指数方面都获得了更好的图像质量，并且在15次迭代下，BP网络的重建速度分别是BPF算法的1.75倍和MLEM算法的29.05倍。结果还表明，时间分辨率越差、示踪剂剂量越小，BP网络的性能就越差，但比BPF或MLEM重建的性能退化得更小。在这项工作中，我们发展了一种解析式的BPF形式的重建，重建滤波操作通过深度网络执行。该方法的运行速度甚至比传统的BPF算法更快，并且可以从TOF-PET中的列表模式数据提供准确的重建，而不需要将数据反弹到正弦图。
The time-of-flight (TOF) information improves signal-to-noise ratio (SNR) for positron emission tomography (PET) imaging. Existing analytical algorithms for TOF PET usually follow a filtered back-projection process on reconstructing images from the sinogram data. This work aims to develop a back-projection-and-filtering-like (BPF-like) algorithm that reconstructs the TOF PET image directly from listmode data rapidly. We extended the 2D conventional non-TOF PET projection model to a TOF case, where projection data are represented as line integrals weighted by the one-dimensional TOF kernel along the projection direction. After deriving the central slice theorem and the TOF back-projection of listmode data, we designed a deep learning network with a modified U-net architecture to perform the spatial filtration (reconstruction filter). The proposed BP-Net method was validated via Monte Carlo simulations of TOF PET listmode data with three different time resolutions for two types of activity phantoms. The network was only trained on the simulated full-dose XCAT dataset and then evaluated on XCAT and Jaszczak data with different time resolutions and dose levels. Reconstructed images show that when compared with the conventional BPF algorithm and the MLEM algorithm proposed for TOF PET, the proposed BP-Net method obtains better image quality in terms of peak signal-to-noise ratio, relative mean square error, and structure similarity index; besides, the reconstruction speed of the BP-Net is 1.75 times faster than BPF and 29.05 times faster than MLEM using 15 iterations. The results also indicate that the performance of the BP-Net degrades with worse time resolutions and lower tracer doses, but degrades less than BPF or MLEM reconstructions. In this work, we developed an analytical-like reconstruction in the form of BPF with the reconstruction filtering operation performed via a deep network. The method runs even faster than the conventional BPF algorithm and provides accurate reconstructions from listmode data in TOF-PET, free of rebinning data to a sinogram.
登录
查看更多内容
影响因子：
10.6
作者：
Parra, L;Barrett, HH
通讯作者：
Barrett, HH
影响因子：
3.5
作者：
Lu, Wenzhuo;Onofrey, John A.;Liu, Chi
通讯作者：
Liu, Chi
影响因子：
10.6
作者：
Shepp, L A;Vardi, Y
通讯作者：
Vardi, Y
DOI：
10.1109/trpms.2020.3004408
发表时间：
2020-06-23
影响因子：
4.4
作者：
Mehranian A;Reader AJ
通讯作者：
Reader AJ
DOI：
10.2967/jnumed.109.063016
发表时间：
2009-08
期刊：
Journal of nuclear medicine : official publication, Society of Nuclear Medicine
影响因子：
--
作者：
Kadrmas DJ;Casey ME;Conti M;Jakoby BW;Lois C;Townsend DW
通讯作者：
Townsend DW