Analysis of Clustering Algorithms for MR Image Segmentation using IQI

This paper presents the analysis of clustering algorithms for medical MR (magnetic resonance) images using IQI (image quality index). The clustering algorithms used are classical C means, fuzzy C means and rough fuzzy C means. With the assistance of the lower and upper approximation of rough sets, the rough fuzzy C means clustering algorithm improves the objective function and further the distribution of membership function for the traditional fuzzy C means clustering. These algorithms are implemented on various medical MR images to detect diseases. The results of various clustering algorithms are compared using IQI index, and it has been found that rough fuzzy C means (RFCM) clustering algorithm fairs better as compared to classical C means and fuzzy C means algorithm respectively.