Reconsidering Explicit Longitudinal Mammography Alignment for Enhanced Breast Cancer Risk Prediction

TL;DR

This paper investigates the optimal strategies for explicit longitudinal mammography alignment, demonstrating that image-level alignment in representation space improves risk prediction accuracy over joint optimization approaches.

Contribution

It provides new insights into where and how to perform explicit alignment in mammography, favoring image-level over representation-level alignment for better risk prediction.

Findings

Image-level alignment yields better deformation quality.

Joint optimization causes a trade-off between alignment and prediction.

Representation-level alignment is less effective than image-level alignment.

Abstract

Regular mammography screening is essential for early breast cancer detection. Deep learning-based risk prediction methods have sparked interest to adjust screening intervals for high-risk groups. While early methods focused only on current mammograms, recent approaches leverage the temporal aspect of screenings to track breast tissue changes over time, requiring spatial alignment across different time points. Two main strategies for this have emerged: explicit feature alignment through deformable registration and implicit learned alignment using techniques like transformers, with the former providing more control. However, the optimal approach for explicit alignment in mammography remains underexplored. In this study, we provide insights into where explicit alignment should occur (input space vs. representation space) and if alignment and risk prediction should be jointly optimized. We demonstrate that jointly learning explicit alignment in representation space while optimizing risk estimation performance, as done in the current state-of-the-art approach, results in a trade-off between alignment quality and predictive performance and show that image-level alignment is superior to representation-level alignment, leading to better deformation field quality and enhanced risk prediction accuracy. The code is available at https://github.com/sot176/Longitudinal_Mammogram_Alignment.git.