Detection of Tumor Region in MRI Images Using Kernel Fuzzy C Means with PSO
R. Sumathi, Saravanan Alagarsamy, P. Nagaraj, R Venkatesh.
Abstract
From the past 10 to 15 years many peoples lost their live due to cancers. Brain cancer, skin cancer are in the leading place for cancer related diseases. MRI imaging is highly preferred for analyzing the detailed information of tumor part in accurate manner. To assist radiologist, an automated method is proposed to detect the abnormal part in efficient manner with the integration of Kernel FCM with PSO. We used image sequences such as T1, T2-Weighted images collected from BRATS, RIDER and Harvard brain datasets for validation. Various metrics like MSE, PSNR sensitivity, specificity and accuracy were validated to ensure the efficiency of abnormal detection. Our hybrid method obtained 57.5 PSNR and processed within 7s duration. It proves its tumor exaction accuracy (97.6%) is superior that other existing approaches.