M&M: Tackling False Positives in Mammography with a Multi-view and Multi-instance Learning Sparse Detector

TL;DR

This paper introduces M&M, a novel mammography detection system that reduces false positives by combining sparse detection, multi-view cross-attention, and multi-instance learning, validated across five datasets.

Contribution

The paper presents a new multi-view, multi-instance learning approach with sparse detectors tailored for mammography, improving false positive reduction and breast-level classification.

Findings

M&M outperforms existing methods in detection accuracy.

Multi-view cross-attention enhances information synthesis.

Multi-instance learning enables training with unannotated images.

Abstract

Deep-learning-based object detection methods show promise for improving screening mammography, but high rates of false positives can hinder their effectiveness in clinical practice. To reduce false positives, we identify three challenges: (1) unlike natural images, a malignant mammogram typically contains only one malignant finding; (2) mammography exams contain two views of each breast, and both views ought to be considered to make a correct assessment; (3) most mammograms are negative and do not contain any findings. In this work, we tackle the three aforementioned challenges by: (1) leveraging Sparse R-CNN and showing that sparse detectors are more appropriate than dense detectors for mammography; (2) including a multi-view cross-attention module to synthesize information from different views; (3) incorporating multi-instance learning (MIL) to train with unannotated images and perform breast-level classification. The resulting model, M&M, is a Multi-view and Multi-instance learning system that can both localize malignant findings and provide breast-level predictions. We validate M&M's detection and classification performance using five mammography datasets. In addition, we demonstrate the effectiveness of each proposed component through comprehensive ablation studies.