A Preliminary Study on Projection Denoising for Low-dose CT Imaging Using Modified Dual-Domain U-net

2020 3rd International Conference on Artificial Intelligence and Big Data (ICAIBD)(2020)

引用 6|浏览9
暂无评分
摘要
Recently, low-dose computed tomography (CT) was considered by many researchers to be a good solution to reduce radiation risks of patients. However, lowering X-ray tube current will make reconstructed images quality be significantly degraded. To improve image quality, in this paper, we proposed a modified dual-domain U-net (MDD-U-net) that combines the projection domain and image domain losses. The proposed MDD-U-net can effectively suppress the projection domain noise and reduce the error in reconstructed images. The simulation experiment results showed that the proposed method effectively reduce the noise in the low-dose CT images while preserved images details information.
更多
查看译文
关键词
low-dose CT,deep learning,image reconstruction,projection denoising
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要