CSAW-M: An Ordinal Classification Dataset for Benchmarking Mammographic Masking of Cancer

TL;DR

This paper introduces CSAW-M, a large annotated mammographic dataset for benchmarking cancer masking, and demonstrates that deep learning models trained on it can better predict difficult-to-detect cancers than traditional density measures.

Contribution

The paper presents CSAW-M, the largest public dataset with expert annotations of mammographic masking, and shows its effectiveness in improving cancer detection predictions.

Findings

Deep learning models trained on CSAW-M outperform density measures in predicting interval and invasive cancers.

CSAW-M is the largest dataset with expert-annotated masking potential for mammography.

Estimated masking levels correlate strongly with difficult-to-detect cancers.

Abstract

Interval and large invasive breast cancers, which are associated with worse prognosis than other cancers, are usually detected at a late stage due to false negative assessments of screening mammograms. The missed screening-time detection is commonly caused by the tumor being obscured by its surrounding breast tissues, a phenomenon called masking. To study and benchmark mammographic masking of cancer, in this work we introduce CSAW-M, the largest public mammographic dataset, collected from over 10,000 individuals and annotated with potential masking. In contrast to the previous approaches which measure breast image density as a proxy, our dataset directly provides annotations of masking potential assessments from five specialists. We also trained deep learning models on CSAW-M to estimate the masking level and showed that the estimated masking is significantly more predictive of screening participants diagnosed with interval and large invasive cancers -- without being explicitly trained for these tasks -- than its breast density counterparts.