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

Abstract

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. Using DDSM as benchmark, our models compare favorably with the current state-of-the-art. We also demonstrate that a whole image model trained on DDSM can be easily transferred to INbreast without using its lesion annotations and using only a small amount of training data. Code availability: https://github.com/lishen/end2end-all-conv

Accepted poster at NIPS 2017 Workshop on Machine Learning for Health (https://ml4health.github.io/2017/)

Cited by in corpus (14)

• Memory-aware curriculum federated learning for breast cancer classification
• Breast Cancer Diagnosis in Two-View Mammography Using End-to-End Trained EfficientNet-Based Convolutional Network
• Is Grad-CAM Explainable in Medical Images?
• Domain generalization in deep learning-based mass detection in mammography: A large-scale multi-center study
• Meta Ordinal Regression Forest for Medical Image Classification with Ordinal Labels
• Multi-task fusion for improving mammography screening data classification
• Weakly and Semi Supervised Detection in Medical Imaging via Deep Dual Branch Net
• Detection of Mild Cognitive Impairment Using Facial Features in Video Conversations
• Evaluating Deep Learning Models for Breast Cancer Classification: A Comparative Study
• Lesion detection in contrast enhanced spectral mammography
• MamT${}^4$: Multi-view Attention Networks for Mammography Cancer Classification
• Augmented Intelligence for Multimodal Virtual Biopsy in Breast Cancer Using Generative Artificial Intelligence
• An XAI Approach to Deep Learning Models in the Detection of DCIS
• Multimodal Breast Lesion Classification Using Cross-Attention Deep Networks