Improving Low-Dose Brain Perfusion Computed Tomography Using 3D Dictionary Learning Based Processing
Improving Low-Dose Brain Perfusion Computed Tomography Using 3D Dictionary Learning Based Processing
复制标题
使用基于 3D 字典学习的处理改进低剂量脑灌注计算机断层扫描
DOI：
10.1166/jmihi.2015.1569
复制
发表时间：
2015-12
影响因子：
--
通讯作者：
Luo Limin
中科院分区：
文献类型：
--
作者：
Shi Luyao;Yin Xindao;Zhang Libo;Yang Benqiang;Zhan Jie;Chen Yang;Shu Huazhong;Luo Limin
虽然与标准剂量扫描相比，低剂量CT灌注（LDCTp）图像的健康风险较低，但往往会受到量子噪声和条纹伪影的严重影响。因此，在本文中，3D字典学习（DL）为基础的处理，提出了提高LDCTp图像质量。通过稀疏表示利用空间和时间连续性的特征信息来提高LDCTp质量。临床数据的实验验证了该方法的良好性能。
Though with lower health risks compared with standard dose scanning, low-dose CT perfusion (LDCTp) images tend to be severely degraded by quantum noise and streak artifacts. Accordingly, in this paper, 3D dictionary learning (DL) based processing is proposed to improve the LDCTp image quality. Feature information on both spatial and temporal continuity is exploited via sparse representation to improve LDCTp quality. Experiments on clinical data validate the good performance of the proposed method.