结合模糊C均值聚类和曲线演化的心脏MRI图像分割

Cardiac MRI Image Segmentation by Unifying Fuzzy C-Means Clustering and Curves Evolving

提出了结合模糊C均值聚类和曲线演化的MRT图像分割算法。由于心脏的变形和血液的流动,MRI图像中出现了弱边界、局部梯度极大值区域、伪影等现象。分析了使用水平集方法分割此类图像时出现的问题,提出了两阶段分割算法:结合先验知识和直方图,对心脏MRI图像进行模糊C均值聚类,再根据聚类的结果定义窄带中像素点的速度函数,通过曲线演化获取左心室的粗边界;然后使用梯度向量流构造另一速度函数对边界进行细化。心脏MRI图像的分割实验证明了算法的有效性。

An MRI image segmentation algorithm is presented by unifying fuzzy C-means clustering and curves evolving. Due to cardiac deformation and blood flowing, weak edges, local gradient maximum regions and artifacts are found in the MRI images. The difficulties are analyzed when the level set method is applied to segment those MRI images. A two-stage algorithm is thus proposed: firstly, incorporating prior knowledge about cardiac MRI and image histogram, the fuzzy C-means clustering is applied and with the clustering results, and the velocity function of the pixels in the narrow band is constructed to search for the rough boundary of left ventricle; secondly, another velocity function based on Gradient Vector Flows (GVF) is designed to locate the edges accurately. The results of the experiments demonstrate the effectiveness of the segmentation algorithm.