模糊C-均值算法在脑MRI图像分割中的应用

Application of fuzzy C-means algorithm tobrain MRI image segmentation

研究了模糊C-均值(fuzzy C-means,FCM)算法在脑MRI图像分割中的应用,介绍了传统FCM算法在图像分割中存在的不足,并从空间连续性与磁场非均匀性校正、基于核函数的FCM算法以及FCM与其他算法联合分割3个方面,阐述了对传统FCM算法的改进,最后指出了改进目标函数和与其他分割算法相结合是今后主要的研究方向。

Application of fuzzy C-means（FCM） algorithm is studied to brain MRI image segmentation. The deficiencies of traditional FCM algorithm are introduced, and the improvement is explored from the aspects of spatial continuity and magnetic field nonuniformity correction, kernel-based FCM algorithm as well as co-segmentation by FCM and other algorithms. The improvement of the traditional FCM algorithm is described, and it＇s pointed out that FCM algorithm has to be enhanced for the objective function and combined with other algorithms.