MRI Image Segmentation using Stationary Wavelet Transform and FCM Algorithm

Image segmentation is one of the vital steps in Image processing. It is a challenging task in segmenting MRI(Magnetic Resonance Imaging) images because these images have no linear features. But MRI images provide high quality when compared to any other imaging techniques, so it is best suited for clinical diagnosis, biomedical research, etc. This paper presents a novel approach for segmenting MRI brain images using Stationary Wavelet Transform (SWT) and Clustering Technique. The clustering technique used here is Fuzzy cmeans (FCM) clustering because it provides better segmentation for medical images. The obtained result using Stationary wavelet transform and Clustering Technique is compared with the existing method. The quality of segmentation is evaluated with deviation ratio as performance measure and the performance comparison for Discrete Wavelet Transform and Stationary Wavelet Transform in segmenting MRI images has been performed and the deviation ratio values are tabulated.