Multi-Head Feature Pyramid Networks for Breast Mass Detection

TL;DR

This paper introduces a multi-head feature pyramid network to enhance breast mass detection in X-ray images, significantly improving accuracy especially for smaller masses, which is crucial for early breast cancer diagnosis.

Contribution

The paper proposes the multi-head feature pyramid module (MHFPN) and a new detection network (MBMDnet) to address unbalanced focus on masses of different sizes during feature fusion.

Findings

6.58% improvement in AP@50 on INbreast dataset

5.4% increase in TPR@50 on INbreast

6-8% improvements in AP@20 on MIAS and BCS-DBT datasets

Abstract

Analysis of X-ray images is one of the main tools to diagnose breast cancer. The ability to quickly and accurately detect the location of masses from the huge amount of image data is the key to reducing the morbidity and mortality of breast cancer. Currently, the main factor limiting the accuracy of breast mass detection is the unequal focus on the mass boxes, leading the network to focus too much on larger masses at the expense of smaller ones. In the paper, we propose the multi-head feature pyramid module (MHFPN) to solve the problem of unbalanced focus of target boxes during feature map fusion and design a multi-head breast mass detection network (MBMDnet). Experimental studies show that, comparing to the SOTA detection baselines, our method improves by 6.58% (in AP@50) and 5.4% (in TPR@50) on the commonly used INbreast dataset, while about 6-8% improvements (in AP@20) are also observed on the public MIAS and BCS-DBT datasets.