Clustering Techniques for Medical Imaging

Abstract

Nowadays medical imaging is becoming one of the popular techniques used to monitor human body to diagnose diseases, detect and treat injuries so that it can be treated. It helps in fetching desired information from the medical images. Clustering techniques in medical imaging is used to assist image based analysis of heterogeneous ailments by creating clusters of given population into homogeneous sub populations which helps in better understanding of the disease within each sub population. In this paper, we have discussed and compared various clustering techniques such as Fuzzy C Means clustering (FCM), Spatial Fuzzy C Means clustering(SFCM), K-Means and Particle Swarm Optimization Incorporative Fuzzy C Means clustering (PSOFCM), Gustafson Kessel (GK) clustering and Density Based Clustering of Applications with Noise (DBSCAN) to detect a tumor in human brain based on various image segmentation parameters. Accuracy of these algorithms is tested using MRI brain image.