Computer-Aided Diagnosis System for Masses Classification (benign - malignant) in Digital Mammograms

In this work a CAD system to distinguish between benign and malignant breast tissue was proposed. Image enhancement is implemented by three techniques; peripheral enhancement and thickness correction, Histogram Equalization (HE) and median filter. Regions of interest (ROIs) containing masses of breast tissue are extracted. Then we extracted 397 features and then performed feature selection using Sequential Forward Selection, Binary Particle Swarm Optimizer and Binary Grey Wolf Optimizer. Finally we used K-Nearest Neighbor (KNN), Support Vector Machine (SVM), Artificial Neural Network (ANN), Ensemble classifier and Multi-Layer Perceptron (MLP) classifiers for classification. The obtained results show acceptable sensitivity and specificity for the system.