Metastatic Cancer Image Binary Classification Based on Resnet Model
Mingrui Wang, Xuhui Gong · 2020
With the rapid development of computers technology, computers have inherent advantages over human in terms of sample storage, data processing, and calculation speed. Computer assisted diagnosis achieves automatical disease detection and classification which is economical. In recent years, using deep learning methods to solve medical image classification has received much attention from researchers due to great performance. In this article, we propose a new metastatic cancer image classification approach based on Resent model. We aim to turn the digital pathology scan detection task into a binary classification task of cancer image. The ResNet model can effectively alleviate the gradient explosion and gradient disappearance problems and train a deeper network by using a skip connection with an identity transformation. The experiments were conducted on the PCam benchmark dataset which is provided by the Kaggle competition platform. The experimental results indicate that our method outperforms the other convolutional neural networks methods such as Vgg16 and Vgg19. Our proposed method has the highest scores on the Auc Roc and the Accuracy measures. Experimental results show that the use of deep learning method based on Resent can significantly improve the performance of metastatic cancer diagnosis.