Kidney Lesion Segmentation in MRI Using Clustering with Salp Swarm Algorithm

Abstract

In this paper, kidney lesion segmentation in MRI using clustering with salp swarm algorithm (SSA) is proposed. The segmentation results of kidney MRI are degraded by the noise and intensity inhomogeneities (IIHs) in MR images. Therefore, at the outset, the MR images are denoised using median filter. Then IIHs are corrected using the max filter-based method. A hard-clustering technique using SSA is developed to segment the MR images. Finally, the lesions are extracted from the segmented MR images. The proposed method is compared with the K-means algorithm using well-known clustering validity measure DB-index. The experimental results demonstrate that the proposed method performs better than the K-means algorithm in the segmentation of kidney lesions in MRI.