Improving localization-based approaches for breast cancer screening exam classification

Thibault Févry, Jason Phang, Nan Wu, S. Gene Kim, Linda Moy, Kyunghyun Cho, Krzysztof J. Geras

Keywords: breast cancer, breast cancer screening exam classification, faster-rcnn, interpretability

TL;DR: Well-tuned localization models perform well at breast cancer screening exam classification

Abstract: We trained and evaluated a localization-based deep CNN for breast cancer screening exam classification on over 200,000 exams (over 1,000,000 images). Our model achieves an AUC of 0.919 in predicting malignancy in patients undergoing breast cancer screening, reducing the error rate of the baseline (Wu et al., 2019a) by 23%. In addition, the models generates bounding boxes for benign and malignant findings, providing interpretable predictions.

Code Of Conduct: I have read and accept the code of conduct.