基于Contourlet域Context模型的磁共振图像去噪方法

磁共振图像（MRI）广泛地应用在医学诊断上，但由于噪声的影响存在，一些重要的信息被淹没。目前，人们把小波应用在磁共振图像的去噪上，但是由于小波方向性不足，常用的一些经典方法门限选择不够恰当，造成处理后，图像纹理特征被弱化，图像边缘变得模糊。本文利用contour-let变换，构建context模型，来实现磁共振图像的去噪。仿真实验结果表明，本方法是有效可用的，与其它方法比较，具有更高的PSNR值和较优的视觉效果。

A Study of MR Image Denoising Method Based on Contourlet Transform with Context Model

Some important information of MIR image may be submerged due to the presence of noise. Therefore, it is important for MR images to be preprocessed before being used for analysis. So far, wavelet transform is often used as the method of MR image de-noising. Wavelet transform has a limitation in directionality, and its threshold is sometimes not appropriately selected in some classical methods. As a result, the texture characteristics of MR images is weakened, and edges of images are blurred after processing. In this paper, we proposed a new method of contourlet transform with context modeling to achieve better denoising of MR images. The experiment results show a good efficiency of the proposed method with a higher PSNR value and better visual effect than other methods.