A Hybrid approach of segmentation in Brain Tumor Data using Fuzzy C-means and Level Set

Allam Balaram,Shaik Abdul Nabi

In worldwide, a leading cause of deaths in human’s causes with number of diseases among popular one is brain tumor. However, treatment for tumor that causes at early stage can affect to raise in the survival rate. The analysis over medical tumor data come up with different techniques among the difficult one is segmentation approach. In the segmentation technique, tumor is extracted from the complete area of brain image and resultant image is forwarded to coming analysis especially object detection or recognition. The work in this paper, focused on designed to a hybrid model which combines fuzzy c-mean and level set methods to improve the performance of segmentation analysis over brain tumor images. Moreover, proposed method carried experiment analysis over real brain images extracted from the data bases like DICOM. The proposed method first carried with the preprocessing of an MRI image and then a Rough FCM algorithm is applied. From the result of RFCM, each data point of image is represented into individual clusters. The next step is applied level set method and retained segmented the tumor boundaries. To estimate the performance of the proposed algorithm over standard methods number of metrics applied which includes precision, sensitivity and specificity. Also observed that proposed method is promising segmentation results compared to standard techniques.

Volume 12 | Issue 6

Pages: 47-56

DOI: 10.5373/JARDCS/V12I6/S20201006