A HYBRID FEATURE EXTRACTION METHOD WITH REGULARIZED EXTREME LEARNING MACHINE FOR BRAIN TUMOR CLASSIFICATION

Abstract

Brain cancer classification is an important step that depends on the physician’s knowledge and experience. An automated tumor classification system is very essential to support radiologists and physicians to identify brain tumors. However, the accuracy of current systems need to be improved for suitable treatments. In this paper, we propose a hybrid feature extraction method with regularized extreme learning machine for developing an accurate brain tumor classification approach. The approach starts by extracting the features from brain images using the hybrid feature extraction method; then, computing the covariance matrix of these features to project them into a new significant set of features using principle component analysis (PCA). Finally, a regularized extreme learning machine (RELM) is used for classifying the type of brain tumor. To evaluate and compare the proposed approach, a set of experiments is conducted on a new public dataset of brain images. Experimental results proved that the approach is more effective compared to the existing state-of-the-art approaches, and the performance in terms of classification accuracy improved from 91.51% to 94.233% for the experiment of random holdout technique.  

Existing System

Markov random field (MRF) model, GLCM techniques.

Proposed System

The main contributions of this work are outlined in the following lines: We propose an automatic brain tumor classification approach to aid the radiologists and physicians in order to identify the type of brain tumors. We propose a novel and effective hybrid feature method referred to as PCA-NGIST, that uses Normalized GIST descriptor with PCA to extract the significant features from brain images without using any kind of image segmentation. It should be noted that, image segmentation methods are affected by changes in illumination and shadowing leading to inaccurate results in brain tumor classification. We use a RELM classifier due to its regularization property to reduce the overfitting problem and its speed for training. We optimize the parameters of the proposed approach using a grid search algorithm. We evaluate the proposed approach using a new public dataset of brain images and compare with the state-of-the-art approaches on the same dataset.  

Architecture

BLOCK DIAGRAM