Brain Tumor Classification using Bat-Ann with BWT Segmentation Technique
The segmentation and classification of infected tumor area from magnetic resonance imaging (MRI) are a major
concern but a tedious and time consuming task achieved by medical specialists, and their accuracy depends on their
knowledge only. Consequently, the utilization of computer aided technology becomes very essential to overcome these
restrictions. In this paper, to increase the performance and diminish the difficulty includes in the image segmentation
method, we have considered Berkeley wavelet transformation (BWT) based tumor segmentation. Moreover, to increase the
accuracy and quality rate, a Bat algorithm based Artificial Neural Network (ANN) classifier by extracting relevant features
from each segmented tissue. The results of proposed technique have been assessed and validated for performance on MRI
images, based on accuracy, sensitivity, specificity, and so on.
Keywords - MRI, BWT, BAT-ANN, Tumor.