Leveraging Transfer Learning for High-Accuracy Breast Cancer Classification from Histopathological Images

Early detection of breast cancer remains an important global health concern. This paper presents a method for classifying breast cancer using histopathological images from the BreakHis dataset at 400× resolution. High-level features capturing malignancy patterns are extracted using VGG19 and DenseNet201. These features are concatenated and fed into an Artificial Neural Network (ANN), achieving 99% accuracy. The results highlight the method’s potential as an effective diagnostic tool for digital pathology.