模糊C-均值聚类法在医学图像分析中的应用

主要针对医学图像提出了基于模糊均值聚类的改进算法和应用.该方法分为3步,第1步是像素的模糊化,通过模糊期望值构造冗余图像;第2步是通过冗余图像和原始图像进行聚类分割;第3步是三维显示.由于利用冗余图像增加了每个像素的特征量,该算法增强了聚类分割的精确度.同时,还给出了应用自行开发的三维医学图像处理与分析系统对多种医学图像(包括CT、螺旋CT和MRI)的处理结果.由于对薄骨和关节接合处骨骼的较好识别,使其重建后的三维模型可以清晰地再现解剖结构,取得了较好的效果.

Application of the Fuzzy C-Means Clustering Algorithm on the Analysis of Medical Images

In this paper, an improved method is proposed based on the Fuzzy C-means method to deal with medical images. This method includes three steps. The first step is the fuzzy pixels process in which a redundant image is built by FEV (fuzzy expectation value). The second step is the procession of FCM (fuzzy C-means clustering) with original images and their redundant images. The last step is the display of 3D model. This algorithm improves the accuracy of clustering as the redundant image increases the feature of pixels. Several results of medical images are exhibited including CT, spiral CT and MRI, which are processed with the 3D MIPA system developed by the authors. Because better segmentation results have been obtained, the system can represent the anatomy structure of bones and the bones in the joint based on recognition and 3D reconstruction.