数据压缩的FCM算法用于人脑MRI图像的分割

针对模糊C均值聚类（Fuzzy C—means clustering，FCM）算法在人脑磁共振成像（Magnetic Resonance Imaging，MRI）图像分割中存在的计算量大、运行时间过长的问题，提出一种加速方法。首先利用边界跟踪法对人脑MRI图像进行预处理，剔除颅骨和肌肉等非脑组织；然后通过数据压缩，即通过对相近的像素进行量化并聚合来减少像素个数；最后用FCM算法对大脑结构进行分割，结果得到脑白质、灰质和脑脊液图像区域。该方法能够使数据量大为减少，从而使FCM在图像分割中更加快速有效。

Data Compressed FCM Algorithm for Image Segmentation of Brain MRI Images

The fuzzy C-means clustering(FCM) algorithm requires a long time to segment images,especially brain magnetic resonance images,due to the large data set to be processed.In this paper,an accelerated method is put forward.Firstly,the non-brain tissues,such as skull and muscle,are moved by edge-tracing method;and then the method of data compression is used to reduce the number of pixels by quantizing and aggregating similar pixels;finally,FCM algorithm is used to segment the brain structure to get the white matter,gray matter and cerebral spinal fluid.Tests show these methods can reduce data amount greatly and make FCM algorithm segment MRI fastly and efficiently.