Two-stage Augmentation for Detecting Malignancy of BI-RADS 3 Lesions in Early Breast Cancer

Abstract

Abstract Objectives In view of inherent attributes of breast BI-RADS 3, benign and malignant lesions are with a subtle difference and the imbalanced ratio (with a very small part of malignancy). The objective of this study is to improve the detection rate of BI-RADS 3 malignant lesions on breast ultrasound (US) images using deep convolution networks. Methods In the study, 1,275 lesions out of 1,096 patients were included from Southwest Hospital (SW) and Tangshan Hospital (TS). In which, 629 lesions, 218 lesions and 428 lesions were utilized for the development dataset, the internal and external testing set. All lesions were confirmed with ground truth of three-year follow-up benign or biopsy benign/malignancy, and each lesion had both B-mode and color Doppler images. We proposed a two-step augmentation method, covering malignancy feature augmentation and data augmentation, and further verified the feasibility of our augmentation method on a dual-branches ResNet50 classification model named Dual-ResNet50. We conducted a comparative analysis between our model and four radiologists in breast imaging diagnosis. Results After malignancy feature and data augmentations, our model achieved a high area under the receiver operating characteristic curve (AUC) of 0.881(95% CI: 0.830-0.921), the sensitivity of 77.8% (14/18), in the SW test set, and an AUC of 0.880 (95% CI: 0.847-0.910), a sensitivity of 71.4% (5/7) in the TS test set. In the comparison set, our model outperformed four radiologists with more than 10-years of diagnosis experience. Our method improved the cancer detection rate of BI-RADS 3 lesions, thereby aiding in a timely adjustment of subsequent treatment for these patients in the early stage. Conclusions The results demonstrated that our proposed augmentation method can help the deep learning (DL) classification model to improve the breast cancer detection rate in BI-RADS 3 lesions.