AN ADVANCED FCM AND LEVEL SET BASED IMAGE SEGMENTATION METHOD FOR MEDICAL APPLICATIONS

Abstract

Images often suffer from noise and intensity inhomogeneity this makes segmentation challenging. Especially in medical images accurate segmentation of the voxels is necessary. They often corrupted by noise and non-uniformity. Fuzzy c means clustering is one of the popular method in medical image segmentation. But this can’t dealt with noise and intensity inhomogeneity. Recently level set based active contour models are also used in medical image analysis. Precise segmentation capability of active contour models make them attractive. Chumming Li’s model is proposed to deal with intensity inhomogeneity by using an energy function based on K-means clustering. In this report we propose a new energy model based on Li’s model. Our addition is twofold, first we introduce a fuzzy factor into the energy function which is somewhat similar to fuzzy c-means clustering in continuous domain and secondly we utilize a special function which will be advantageous for segmenting noisy image. The proposed method can dealt with intensity inhomogeneity and noise as well. Even in the presence on noise it can result in smooth boundaries. The proposed method is verified on different MRI image which contain noise and intensity inhomogeneity and also on some natural images as well as synthetic images. The results shows that the proposed method is showing improved performance when compared with the state of the art techniques when dealing with images containing inhomogeneity and noise.