期刊
IEEE TRANSACTIONS ON RADIATION AND PLASMA MEDICAL SCIENCES
卷 5, 期 1, 页码 108-119出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TRPMS.2020.2989634
关键词
Computed tomography (CT) reconstruction; denoising autoencoder (AE) network; multichannel; proximal gradient descent; sparse-view; variable augmentation
资金
- National Natural Science Foundation of China [61871206, 61661031, 61671312]
- Basic Research Program of Shenzhen [JCYJ20150831154213680]
The paper introduces a robust enhancement mechanism for sparse-view computed tomography reconstruction, using denoising autoencoding prior method, which can substantially improve the reconstruction quality.
In X-ray computed tomography, radiation doses are harmful but can be significantly reduced by intuitively decreasing the number of projections. However, less projection views usually lead to low-resolution images. To address this issue, we propose a robust and enhanced mechanism on the basis of denoising autoencoding prior, or robust EDAEP (REDAEP) for sparse-view computed tomography reconstruction. REDAEP can substantially improve the reconstruction quality with two novel contributions. First, by employing the variable augmentation technique, REDAEP learns higher-dimensional network with three-channel image and proceeds to the single-channel image reconstruction. Second, REDAEP replaces the L-2 regression loss function with a more robust L-p (0 < p < 2) regression to preserve more texture details. The empirical results demonstrate that REDAEP can achieve better performance than state-of-the-arts, in terms of quantitative measures and subjective visual quality.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据