Asian Journal of Information Technology

Year: 2016
Volume: 15
Issue: 1
Page No. 142 - 155

Hybrid Swarm Intelligence Based FA with Modified Levenberg Marquardt Classifier for Detection of Brain Tumors Through Brain MRI Images

Authors : A. Shenbagarajan, V. Ramalingam, C. Balasubramanian and S. Palanivel

Abstract: Most commonly occurring causable disease among human beings is the Brain Tumor, subsequently early discover of brain tumor is important. In order to get over discover an issues of tumor from brain employing MRI images, an efficient classification method for classifying the brain MRI as Tumourous and non-Tumourous classes by making use of hybrid classifier is proposed. There are three critical early stages of brain MRI image analysis prior to the classification process. In the first step is the preprocessing stage which uses the Adaptive Median Filtering (AMF) method for noise elimination from brain MRI, then follows the segmentation process on these enhanced images on the basis of the regions of the images making use of Active contour Model (ACM) with extracted surface feature. The surface feature extraction is done by employing Kernelised Fuzzy-C-Means (KFCM), later from these segmented images, two most vital features are extracted based on the image edges, which are texture and shape. Texture features and shape features are extracted by utilizing hybrid wavelet transform and Sobel, Canny methods. Then, the most significant feature for classification process is chosen using Principal Component Analysis (PCA), finally, hybrid Firefly Algorithm-Modified Levenberg Marquardt (FAMLM) classifier used for classification. The experimental results of this proposed technique demonstrates that the efficiency of the hybrid classifier outperforms than that of the existing classifier.

How to cite this article:

A. Shenbagarajan, V. Ramalingam, C. Balasubramanian and S. Palanivel, 2016. Hybrid Swarm Intelligence Based FA with Modified Levenberg Marquardt Classifier for Detection of Brain Tumors Through Brain MRI Images. Asian Journal of Information Technology, 15: 142-155.

Design and power by Medwell Web Development Team. © Medwell Publishing 2022 All Rights Reserved