磁共振图像中非均匀场的校正

Correction of the non-uniformity field in MR images

缓慢变化的非均匀场使磁共振图像的局部统计特性发生变化 ,不同生理组织的亮度交叠分布 ,使磁共振图像的分割比其他医学图像分割困难的多。磁共振图像中的非均匀场是磁共振图像自动分割的主要障碍。人们提出了众多的磁共振图像非均匀场的校正方法 ,其中有传统的图像处理方法 ,如图像模糊、平滑、滤波 ,也有新的方法 ,如基于分割的方法 ,基于直方图的方法等。

Magnetic resonance images are always corrupted by a slow varying field known as the non uniformity field.The non uniformity modifies the local statistical characteristics of tissues to cause the distribution of signal intensities associated with these tissues to overlap significantly.It makes the intensity based segmentation in MR images far more difficult than in any other medical images.So it is the major obstacle of automatic segmentation in MR images.Many correction approaches have been proposed,including the conventional image processing methods such as image blurring,smoothing,filtering and relatively new segmentation based method and histogram based method.This paper summarizes and discusses these correting methods.