摘要
针对低剂量CT重建图像受噪声污染严重的问题,提出了一种基于分数阶各项异性扩散和小波的MLEM低剂量CT重建算法.首先,采用最大似然期望最大化(Maximum Likelihood Expectation Maximization,MLEM)算法重建低剂量投影数据.然后,将小波变换应用于图像,使得图像的低频系数部分集中了主要信息,而高频系数部分集中了边缘和噪声.最后,在低频系数部分进行基于差分的分数阶各项异性扩散,在高频系数部分进行软阈值处理.采用Shepp-Logan和Hot-Cold模型进行低剂量CT图像重建仿真,实验结果表明,本文算法在降低噪声和保持图像细节方面都优于其他算法.
Concerning the noise problem about reconstructed images of low-dose CT, a MLEM low-dose CT reconstruction algorithm based on fractional-order anisotropic diffusion and wavelet was proposed.Firstly, the MLEM algorithm was used to do reconstruction of the low-dose projection data.Secondly, by the wavelet transformation, the main information of image was concentrated in the low frequency coefficients, and the edge and noise was stored in the high frequency coefficients.Finally, the low frequency coefficients of the image were filtered by fractional-order anisotropic diffusion based on difference, and the high frequency coefficients were processed by the soft threshold function.Shepp-Logan and Hot-Cold model were used for low-dose CT reconstruction simulation.The experimental results demonstrate that compared with other methods, the proposed algorithm achieves superior performance in terms of both noise suppression and detail preservation.
出处
《中北大学学报（自然科学版）》
北大核心
2017年第3期348-353,359,共7页
Journal of North University of China(Natural Science Edition)
基金
山西省重点研发计划资助项目(201603D121012)
关键词
分数阶微分
各项异性扩散
小波变换
低剂量CT
图像重建
fractional-order differentiation
anisotropic diffusion
wavelet transform
image reconstruction
