Breast Cancer Classification from Histopathological Images Using Transfer Learning and Deep Neural Networks

Early diagnosis of breast cancer is the most reliable and practical approach to managing cancer. Computer-aided detection or computer-aided diagnosis is one of the software technology designed to assist doctors in detecting or diagnose cancer and reduce mortality via using the medical image analysis with less time. Recently, medical image analysis used Convolution Neural Networks to evaluate a vast number of data to detect cancer cells or image classification. In this thesis, we implemented transfer learning from pre-trained deep neural networks ResNet18, Inception-V3Net, and ShuffleNet in terms of binary classification and multiclass classification for breast cancer from histopathological images. We use transfer learning with the fine-tuned network results in much faster and less complicated training than a training network with randomly initialized weights from scratch. Our approach is applied to image-based breast cancer classification using histopathological images from public dataset BreakHis. The highest average accuracy achieved for binary classification of benign or malignant cases was 97.11% for ResNet 18, followed by 96.78% for ShuffleNet and 95.65% for Inception-V3Net. In terms of the multiclass classification of eight cancer classes, the average accuracies for pre-trained networks are as follows. ResNet18 achieved 94.17%, Inception-V3Net 92.76% and ShuffleNet 92.27%.