Fang Shiting, Wang Huafeng, Liu Yueliang, Zhang Minghui, Yang Wei, Feng Qianjin, Chen Wufan, Zhang Yu
School of Biomedical Engineering, Southern Medical University, Guangzhou, Guangdong 510515, People's Republic of China. Guangdong Provincial Key Laboratory of Medical Image Processing, Southern Medical University, Guangzhou, Guangdong 510515, People's Republic of China.
Phys Med Biol. 2017 Oct 3;62(20):7925-7937. doi: 10.1088/1361-6560/aa8a48.
Lung 4D computed tomography (4D-CT), which is a time-resolved CT data acquisition, performs an important role in explicitly including respiratory motion in treatment planning and delivery. However, the radiation dose is usually reduced at the expense of inter-slice spatial resolution to minimize radiation-related health risk. Therefore, resolution enhancement along the superior-inferior direction is necessary. In this paper, a super-resolution (SR) reconstruction method based on a patch low-rank matrix reconstruction is proposed to improve the resolution of lung 4D-CT images. Specifically, a low-rank matrix related to every patch is constructed by using a patch searching strategy. Thereafter, the singular value shrinkage is employed to recover the high-resolution patch under the constraints of the image degradation model. The output high-resolution patches are finally assembled to output the entire image. This method is extensively evaluated using two public data sets. Quantitative analysis shows that the proposed algorithm decreases the root mean square error by 9.7%-33.4% and the edge width by 11.4%-24.3%, relative to linear interpolation, back projection (BP) and Zhang et al's algorithm. A new algorithm has been developed to improve the resolution of 4D-CT. In all experiments, the proposed method outperforms various interpolation methods, as well as BP and Zhang et al's method, thus indicating the effectivity and competitiveness of the proposed algorithm.
肺部四维计算机断层扫描（4D-CT）是一种时间分辨CT数据采集技术，在明确将呼吸运动纳入治疗计划和实施过程中发挥着重要作用。然而，为了将辐射相关的健康风险降至最低，通常会以牺牲层间空间分辨率为代价来降低辐射剂量。因此，沿上下方向提高分辨率是必要的。本文提出了一种基于块低秩矩阵重建的超分辨率（SR）重建方法，以提高肺部4D-CT图像的分辨率。具体而言，通过使用块搜索策略构建与每个块相关的低秩矩阵。此后，在图像退化模型的约束下，采用奇异值收缩来恢复高分辨率块。最后将输出的高分辨率块组装起来以输出整个图像。使用两个公共数据集对该方法进行了广泛评估。定量分析表明，相对于线性插值、反投影（BP）和Zhang等人的算法，所提出的算法将均方根误差降低了9.7%-33.4%，边缘宽度降低了11.4%-24.3%。已开发出一种新算法来提高4D-CT的分辨率。在所有实验中，所提出的方法均优于各种插值方法以及BP和Zhang等人的方法，从而表明了所提算法的有效性和竞争力。
