STA-Risk: A Deep Dive of Spatio-Temporal Asymmetries for Breast Cancer Risk Prediction

TL;DR

STA-Risk is a Transformer-based model that captures detailed spatial and temporal asymmetries in mammogram images to improve breast cancer risk prediction over existing models, demonstrating superior performance on multiple datasets.

Contribution

The paper introduces a novel Transformer model with side and temporal encoding for analyzing spatial-temporal asymmetries in mammograms, advancing breast cancer risk prediction methods.

Findings

Outperforms four state-of-the-art models in risk prediction accuracy.

Effectively captures fine-grained spatial and temporal breast tissue changes.

Achieves superior results on two independent datasets.

Abstract

Predicting the risk of developing breast cancer is an important clinical tool to guide early intervention and tailoring personalized screening strategies. Early risk models have limited performance and recently machine learning-based analysis of mammogram images showed encouraging risk prediction effects. These models however are limited to the use of a single exam or tend to overlook nuanced breast tissue evolvement in spatial and temporal details of longitudinal imaging exams that are indicative of breast cancer risk. In this paper, we propose STA-Risk (Spatial and Temporal Asymmetry-based Risk Prediction), a novel Transformer-based model that captures fine-grained mammographic imaging evolution simultaneously from bilateral and longitudinal asymmetries for breast cancer risk prediction. STA-Risk is innovative by the side encoding and temporal encoding to learn spatial-temporal asymmetries, regulated by a customized asymmetry loss. We performed extensive experiments with two independent mammogram datasets and achieved superior performance than four representative SOTA models for 1- to 5-year future risk prediction. Source codes will be released upon publishing of the paper.