Multi-View Mammography Breast Density Classification by SE-ResNet101
WeiChao Yuan, Xufeng Yao, Zezhou Hao, Yanling Yang · 2024
Breast cancer (BC) is among the most frequently tumors in women. Early screening is important in the treatment of BC. Breast density is considered to be closely related to the risk of BC, which has attracted widespread attention from researchers. This study proposes a multi-view SE-ResNet101 model for breast density classification on mammograms from four different views. Compared to original model, proposed model achieved robust classification performance with an accuracy of 0.796, an F1 score of 0.796. The results demonstrated that the model achieved higher accuracy and reliability in breast density classification and was able to provide more effective support for breast cancer assessment.