改进的非局部均值低剂量CT统计迭代重建算法

摘要

为解决低剂量计算机断层扫描(computed tomography,CT)重建图像时产生严重退化的问题,提出一种改进的非局部均值低剂量CT统计迭代重建算法.采用高斯滤波函数对含噪图像进行滤波,利用改进的非局部均值(non-local means,NLM)降噪模型做进一步降噪处理,通过空间邻近度因子和空间变化的滤波参数改进权值函数,得到新的降噪模型,把该模型应用到惩罚加权最小二乘(penalized weighted least square,PWLS)重建算法中,以期达到噪声抑制和边缘保持的良好效果.实验结果表明,该算法的重建图像可有效去除噪声,保护图像的边缘信息和细小结构.%To solve the serious degradation problems of the low-dose computed tomography (CT) reconstruction images,a statistical iterative reconstruction algorithm for low-dose CT based on proposed non-local means (NLM) was proposed.Gaussian filter function was used to filter the noisy image,and the improved NLM noise reduction model was used to reduce noise further.Filtering parameters of spatial proximity factor and spatial variation were used to improve weighting function.New noise reduction model was obtained.The new model was applied to reconstruction algorithm of the penalized weighted least square (PWLS),to obtain good effects on noise suppression and edge preservation.Experimental results show that the proposed algorithm can not only suppress noises,but also well preserve the edge information and fine structures of the images.