FCM与KFCM-II算法在医学MRI图像分割中的应用研究

医学图像分割在医学图像处理,尤其是临床诊断的MRI图像分析中起着重要的作用.由于医学成像过程中存在着各种退化因素,当前各种分割算法仍难以很好地满足高层应用的需求.为解决这一问题,FCM (Fuzzy C-means clustering)算法和它的核方法版本KFCM(Kernel-based FCM)可以应用于图像分割以取得更好的性能表现.对FCM和KFCM-II算法应用于MRI图像分割进行了比较,讨论了在这两种算法中应用灰度有偏场纠正的效果.实验结果表明,在FCM和KFCM-II中采用有偏场模型可以取得更好的分割性能.

FCM and KFCM-II Algorithm with Application in Medical MRI Image Segmentation

Medical image segmentation plays an important role in medical image processing, especially in MRI (Magnetic Resonance Imaging) image analysis for clinical diagnosis. Due to a variety of degradation factors in medical imaging process, the existed segmentation algorithms are still hard to satisfyingly meet the need of higher layer applications. For tackling this problem, FCM (Fuzzy C-means Clustering) and its kernel-based version KFCM (Kernel-based FCM) can be applied to image segmentation and achieve better performance. A comparison of FCM and KFCM-II algorithm with application in MRI image segmentation has been presented, and a discussion on the effect of intensity bias field correction in KFCM-II algorithm given. The experiments show that better segmentation performance can be achieved by adopting bias field model to FCM and KFCM algorithms.