Breast Cancer Medical Image Analysis Based on Transfer Learning Model

Breast cancer becomes one of the common cancer among women. The computer-aided diagnosis is a very available technology, and it becomes an inevitable trend of modern medicine. There is litter data in the medical field which is not enough to support training model. The paper aims to solve the problem in the case of insufficient data volume, and a transfer learning model combined with Convolutional Neural Network (CNN) is proposed to achieve the goal. This model has three innovations. The first one takes Xavier method to initialize parameter, which can make the training process more stable. The second innovation takes dropout method to discard some network nodes randomly, which can reduce overfitting problem. The third one adds two convolutional layers and one max pooling layer before final fully connected layer. The experimental results have shown that this strategy is suitable for the problem in this paper. The paper indicates that the transfer learning model is an effective method with small-scale data, and it can be combined with deep learning algorithms.