Brain Disorder Detection

This study introduces a robust computer-aided diagnosis (CAD) system for detecting multiple brain abnormalities, aimed at assisting physicians in diagnosing and treating brain diseases. The system uses a combination of wavelet sub-band energy, textural, and intensity features from MR brain images, which are ranked by the Wilcoxon test and classified using a backpropagation neural network with Bayesian regulation. The system achieved 100% accuracy in classifying 90 MR images into 18 classes and 97.81% accuracy in classifying 310 MR images into 6 classes, demonstrating its potential for clinical application in identifying multiple brain disorders.