Machine Learning

End-to-end Training for Whole Image Breast Cancer Diagnosis using An All Convolutional Design

Tagged: , ,

This topic contains 0 replies, has 1 voice, and was last updated by  arXiv 1 year, 3 months ago.


  • arXiv
    5 pts

    End-to-end Training for Whole Image Breast Cancer Diagnosis using An All Convolutional Design

    We develop an end-to-end training algorithm for whole-image breast cancer diagnosis based on mammograms. It requires lesion annotations only at the first stage of training. After that, a whole image classifier can be trained using only image level labels. This greatly reduced the reliance on lesion annotations. Our approach is implemented using an all convolutional design that is simple yet provides superior performance in comparison with the previous methods. On DDSM, our best single-model achieves a per-image AUC score of 0.88 and three-model averaging increases the score to 0.91. On INbreast, our best single-model achieves a per-image AUC score of 0.96. Based on the same data, our models beat the top-performing teams method from a recent breast cancer diagnosis competition. We also demonstrate that a whole image model trained on DDSM can be easily transferred to INbreast using only a small amount of training data. Code and model availability: https://github.com/lishen/end2end-all-conv

    End-to-end Training for Whole Image Breast Cancer Diagnosis using An All Convolutional Design
    by Li Shen
    https://arxiv.org/pdf/1708.09427v2.pdf

You must be logged in to reply to this topic.