Classifying Mammographic Breast Density by Residual Learning

Mammographic breast density, a parameter used to describe the proportion of breast tissue fibrosis, is widely adopted as an evaluation characteristic of the likelihood of breast cancer incidence. Existing methods of breast density classification either requires steps of manual operations or achieves only moderate classification accuracies due to the limited model capacity. In this study, we present a radiomics approach based on residual learning for the classification of mammographic breast densities. Different from those established approaches, our method possesses several encouraging properties including being almost fully automatic, possessing big model capacity, and having high flexibility. As a result, it can obtain outstanding classification results without the necessity of result compensation using mammographs taken from different views. The proposed method was instantiated with the INbreast dataset and classification accuracies of 92.6% and 96.8% were obtained for the four BI-RADS (Breast Imaging and Reporting Data System) category task and the two BI-RADS category task, respectively. Both values are significantly higher than the classification results of the current state-of-the-art methods, including the eight-layer convolutional neural network and the high throughput-derived multilayer visual representations. The superior performances achieved with its encouraging properties indicate that our method has a great potential to be applied as a computer-aided diagnosis tool.