论文标题
通过设计训练过程来改善乳房X线摄影恶性细分
Improving Mammography Malignancy Segmentation by Designing the Training Process
论文作者
论文摘要
我们从事乳房成像的恶性细分任务，同时专注于训练过程而不是网络复杂性。我们设计了一个基于修改的U-NET的培训过程，通过使用良性和恶性数据进行培训来增加整体细分性能。我们的方法仅利用少量注释的数据，而依赖于从自我监管的重建任务中转移学习，并有利于解释性。
We work on the breast imaging malignancy segmentation task while focusing on the training process instead of network complexity. We designed a training process based on a modified U-Net, increasing the overall segmentation performances by using both, benign and malignant data for training. Our approach makes use of only a small amount of annotated data and relies on transfer learning from a self-supervised reconstruction task, and favors explainability.
