Photometric Transformer Networks and Label Adjustment for Breast Density Prediction
This work addresses the challenge of accurate and consistent breast density grading for medical imaging, particularly in multi-site data, but it is incremental as it builds on existing deep learning methods.
The paper tackled the problem of breast density prediction from digital mammograms, which is sensitive to normalization settings and reader variability, by proposing photometric transformer networks and label distillation techniques, resulting in significant performance improvement compared to previous approaches.
Grading breast density is highly sensitive to normalization settings of digital mammogram as the density is tightly correlated with the distribution of pixel intensity. Also, the grade varies with readers due to uncertain grading criteria. These issues are inherent in the density assessment of digital mammography. They are problematic when designing a computer-aided prediction model for breast density and become worse if the data comes from multiple sites. In this paper, we proposed two novel deep learning techniques for breast density prediction: 1) photometric transformation which adaptively normalizes the input mammograms, and 2) label distillation which adjusts the label by using its output prediction. The photometric transformer network predicts optimal parameters for photometric transformation on the fly, learned jointly with the main prediction network. The label distillation, a type of pseudo-label techniques, is intended to mitigate the grading variation. We experimentally showed that the proposed methods are beneficial in terms of breast density prediction, resulting in significant performance improvement compared to various previous approaches.