Performance Analysis for Efficient Brain Tumor Segmentation by using Clustering Algorithm
K. Rajesh Babu, P. V. Naganjaneyulu, K. Satya Prasad · Indian Journal of Science and Technology · 2017
Objective: Normally MRI scan or CT helps to view the biology of brain. The segmentation methods are used to identify the tumor size and location. Methods/Analysis: Some of the segmentation methods are the Histogram-based segmentation and the Region-based segmentation (e.g.: Edge Detection method) which have the drawbacks in detection of size of the tumor and region. We are using the clustering based segmentation algorithms in this project. The run time and efficiency are the parameters used for comparison. Findings: These clustering algorithms like K-means, Fuzzy C and Pillar means are compared to each other for better performance by calculating the run time and efficiency of algorithms. This attempt improves the efficiency and computing time. Application/Improvements: It may help pathologists to identify the exact size and region easily. Keywords: Fuzzy C, K-means, Pathologists, Pillar means, Segmentation