Brain Tumor Image Segmentation Based on Spatial Fuzzy Kernel Clustering

Teng-da ZHANG,Xiao-qi LV,Xiao-ying REN,Yu GU,Ming ZHANG

DOI: https://doi.org/10.14107/j.cnki.kzgc.151030

2017-01-01

Abstract:Aiming at the problem of the inaccuracy of spilting due to the sensitiveness to noise and no consideration of the spatial information when segmenting the brain tumor images by the traditional FCM algorithm,a new segmentation method based on spatial fuzzy kemel clustering for brain tumor MR image is proposed.Firstly,the kernel function is introduced into the algorithm which transforms the data from the low dimensional feature space into the high dimensional feature space,and the difference of category features is enhanced.Then,the KFCM algorithm is improved based on the difference of the image target area gradient.The spatial information of images is introduced into the KFCM,and the sum of the target area variance and the sum of the borderline gradient mean reciprocal are used as the new objective function.Finally,the sample characteristics of brain tumor images are optimized iteratively to achieve the accurate segmentation of brain tumor.The precision and robustness of the algorithm have been improved by increasing the space information of the image.The experimental results show that the improved algorithm has high degree of differentiation in different categories and strong noise suppression capability for brain tumor MR images,and achieves good segmentation effect.