Estimating dual-energy CT imaging from single-energy CT data with material decomposition convolutional neural network ?
Estimating dual-energy CT imaging from single-energy CT data with material decomposition convolutional neural network ?
复制标题
使用材料分解卷积神经网络从单能 CT 数据估计双能 CT 成像 ?
DOI：
10.1016/j.media.2021.102001
复制
发表时间：
2021-02-26
影响因子：
10.9
通讯作者：
Xing, Lei
中科院分区：
文献类型：
--
作者：
Lyu, Tianling;Zhao, Wei;Xing, Lei
双能计算机断层扫描（DECT）因其提供材料特定信息的巨大潜力而对临床实践具有重要意义。然而，DECT 扫描仪通常比标准单能 CT (SECT) 扫描仪更昂贵，因此不太适合不发达地区使用。在本文中，我们表明，深度学习模型可以利用标准 DECT 图像之间的能量域相关性和解剖一致性，从完全采样的低能量数据和单视图高能量数据中提供高性能 DECT 成像。我们通过两个独立队列（第一个队列包括来自 22 名患者的 5753 个图像切片的对比增强 DECT 扫描，第二个队列包括来自其他 22 名患者的 2463 个图像切片的未注射对比剂的能谱 CT 扫描）证明了该方法的可行性，并展示了其在 DECT 应用中的优越性能。基于深度学习的方法可用于进一步显着降低当前优质 DECT 扫描仪的辐射剂量，并有可能简化 DECT 成像系统的硬件并使用标准 SECT 扫描仪实现 DECT 成像。(c) 2021 Elsevier B.V. 保留所有权利。
Dual-energy computed tomography (DECT) is of great significance for clinical practice due to its huge potential to provide material-specific information. However, DECT scanners are usually more expensive than standard single-energy CT (SECT) scanners and thus are less accessible to undeveloped regions. In this paper, we show that the energy-domain correlation and anatomical consistency between standard DECT images can be harnessed by a deep learning model to provide high-performance DECT imaging from fully-sampled low-energy data together with single-view high-energy data. We demonstrate the feasibility of the approach with two independent cohorts (the first cohort including contrast-enhanced DECT scans of 5753 image slices from 22 patients and the second cohort including spectral CT scans without contrast injection of 2463 image slices from other 22 patients) and show its superior performance on DECT applications. The deep-learning-based approach could be useful to further significantly reduce the radiation dose of current premium DECT scanners and has the potential to simplify the hardware of DECT imaging systems and to enable DECT imaging using standard SECT scanners.(c) 2021 Elsevier B.V. All rights reserved.