Variants of Fuzzy C-Means on MRI Modality for Cancer Image Archives

The segmentation, identification and mining of contaminated tumor region from MRI images is a primary concern in medical image analysis. However, it is a monotonous and time-consuming process done by radiologists or clinical experts, and its precision is subjected to their expertise. To overcome these constraints, the use of supporting technology turns out to be very important. In this study, performance improvement and reduction of the complexity involved in the segmentation of medical images is been focused. We have investigated cancer cells using various fuzzy c-means methods and its different variants for Breast, Brain, Liver and Prostate dataset. The empirical outcomes of the various methods have been tested and corroborated on MRI for efficiency and quality analysis based on four well-known cluster validity indices. The results are very encouraging and robust in nature.