模糊K-均值聚类算法及其在磁共振颅脑图像分割中的应用研究

目的　介绍一种动态模糊聚类算法并利用该算法对磁共振图像进行分割研究。方法　首先对磁共振颅脑图像进行预处理去掉颅骨和肌肉等非脑组织，只保留大脑组织，然后利用模糊Ｋ－ 均值聚类算法计算脑白质、脑灰质和脑脊液的模糊类属函数。结果　模糊Ｋ－ 均值聚类算法能很好地分割出磁共振颅脑图像中的灰质、白质和脑脊液。结论　利用模糊Ｋ－ 均值聚类算法分割磁共振颅脑图像能获得较好的分割效果。

Fuzzy K means Clustering Algorithm and It's Application Study in Segmentation of MR Head Images

Objective To introduce a dynamic fuzzy clustering algorithm and use it to do the study of segmentation of the brain in MRI.Methods At first,romoving the nonbrain tissues such as the skull and muscle from the MR head image with image preprocessing,and then using fuzzy K means clustering algorithm to calculate the fuzzy membership functions for white matter,gray matter and cerebrospinal fluid (CSF).Results Fuzzy K means clustering algorithm can segment white matter,gray matter and CSF better from the MR head images.Conclusion Using fuzzy k means clustering algorithm to segment MR head images can get better segmentation result. [