Transfer Learning-Based Diagnosis of Breast Cancer

Among cancer types affecting women worldwide, breast cancer ranks as one of the leading causes of death, and early detection through screening is vital for improving survival outcomes. This study presents a deep learning-based method for classifying mammogram images into benign and malignant categories using the MIAS dataset. The proposed approach includes preprocessing mammograms, extracting Region of Interest patches, and classifying them using Convolutional Neural Networks with Transfer Learning and data augmentation to address data scarcity. The evaluation results demonstrate that the framework achieves promising classification performance.