Breast Cancer Classification from Histopathological Images using Convolutional Neural Network
Mohiuddin Ahmed, Md. Rabiul Islam · 2021
Breast cancer is one of the leading causes of mortality among women, and it has a severe impact on global public health. Breast cancer diagnosis in its early stages can save thousands of lives every year. The clinical diagnosis is based mostly on subjective analysis of the breast histopathological images and requires a considerable lot of expertise from the pathologists. Computer-aided diagnosis (CAD) system can be used to provide a second opinion to make the breast cancer diagnosis error-free as much as possible. BreakHis, a publicly available breast histopathological image dataset, is used in this study. In this research, a deep convolutional neural network with EfficientNet-B0 as the feature extractor is used to classify the histopathological images of the breast tissues. In terms of accuracy, precision, recall, F-measure, and Area Under Curve (AUC), this model beat prior research by a considerable margin.