Spatial Multi-Task Learning for Breast Cancer Molecular Subtype Prediction from Single-Phase DCE-MRI

2026-01-11

arXiv.org

AI Summary

This study addresses the challenge of non-invasively predicting breast cancer molecular subtypes using only routine single-phase DCE-MRI by proposing a spatial multi-task learning framework that simultaneously predicts ER, PR, HER2 status, and Ki-67 index. The method integrates multi-scale spatial attention with a region-of-interest weighting mechanism to effectively model intra- and peritumoral heterogeneity, while leveraging multi-task learning to uncover shared representations and intrinsic relationships among biomarkers. Evaluated on a cohort of 960 patients, the model achieved AUCs of 0.893, 0.824, and 0.857 for ER, PR, and HER2 classification, respectively, and a mean absolute error of 8.2% for Ki-67 regression—significantly outperforming both radiomics-based and single-task deep learning approaches. This work represents the first demonstration of joint multi-biomarker prediction from single-phase DCE-MRI.

Technology Category

• Machine Learning: Multi-instance/Multi-view Learning
• Computer Vision: Multi-modal Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Accurate molecular subtype classification is essential for personalized breast cancer treatment, yet conventional immunohistochemical analysis relies on invasive biopsies and is prone to sampling bias. Although dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) enables non-invasive tumor characterization, clinical workflows typically acquire only single-phase post-contrast images to reduce scan time and contrast agent dose. In this study, we propose a spatial multi-task learning framework for breast cancer molecular subtype prediction from clinically practical single-phase DCE-MRI. The framework simultaneously predicts estrogen receptor (ER), progesterone receptor (PR), human epidermal growth factor receptor 2 (HER2) status, and the Ki-67 proliferation index -- biomarkers that collectively define molecular subtypes. The architecture integrates a deep feature extraction network with multi-scale spatial attention to capture intratumoral and peritumoral characteristics, together with a region-of-interest weighting module that emphasizes the tumor core, rim, and surrounding tissue. Multi-task learning exploits biological correlations among biomarkers through shared representations with task-specific prediction branches. Experiments on a dataset of 960 cases (886 internal cases split 7:1:2 for training/validation/testing, and 74 external cases evaluated via five-fold cross-validation) demonstrate that the proposed method achieves an AUC of 0.893, 0.824, and 0.857 for ER, PR, and HER2 classification, respectively, and a mean absolute error of 8.2\% for Ki-67 regression, significantly outperforming radiomics and single-task deep learning baselines. These results indicate the feasibility of accurate, non-invasive molecular subtype prediction using standard imaging protocols.

Problem

• breast cancer
• molecular subtype prediction
• single-phase DCE-MRI
• non-invasive diagnosis
• biomarker classification

Innovation

• spatial multi-task learning
• single-phase DCE-MRI
• multi-scale spatial attention
• region-of-interest weighting
• molecular subtype prediction