Attention-Enhanced Deep Learning Ensemble for Breast Density Classification in Mammography

2025-07-08

AI Summary

Breast density (BI-RADS categories C/D) is a critical biomarker for breast cancer risk prediction and imaging detection; however, manual assessment suffers from substantial inter-observer variability and low efficiency. To address this, we propose a clinically standardized automated binary classification method—distinguishing BI-RADS A/B from C/D—based on a channel-attention-enhanced deep learning architecture trained with a novel weighted Focal-Dice hybrid loss function. To mitigate class imbalance and improve robustness, we further introduce a multi-model ensemble voting strategy. Our framework integrates ResNet18/50, EfficientNet-B0, and DenseNet121 backbones, augmented by CLAHE preprocessing and comprehensive data augmentation. Evaluated on the VinDr-Mammo dataset, the proposed method achieves an AUC of 0.963 and an F1-score of 0.952—significantly outperforming individual baseline models—demonstrating high accuracy, strong generalizability, and promising clinical deployability.

Technology Category

Computer Vision: Medical and Biological ImagingMachine Learning: Ensemble MethodsReasoning under Uncertainty: Graphical Models

Application Category

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

Abstract

Breast density assessment is a crucial component of mammographic interpretation, with high breast density (BI-RADS categories C and D) representing both a significant risk factor for developing breast cancer and a technical challenge for tumor detection. This study proposes an automated deep learning system for robust binary classification of breast density (low: A/B vs. high: C/D) using the VinDr-Mammo dataset. We implemented and compared four advanced convolutional neural networks: ResNet18, ResNet50, EfficientNet-B0, and DenseNet121, each enhanced with channel attention mechanisms. To address the inherent class imbalance, we developed a novel Combined Focal Label Smoothing Loss function that integrates focal loss, label smoothing, and class-balanced weighting. Our preprocessing pipeline incorporated advanced techniques, including contrast-limited adaptive histogram equalization (CLAHE) and comprehensive data augmentation. The individual models were combined through an optimized ensemble voting approach, achieving superior performance (AUC: 0.963, F1-score: 0.952) compared to any single model. This system demonstrates significant potential to standardize density assessments in clinical practice, potentially improving screening efficiency and early cancer detection rates while reducing inter-observer variability among radiologists.

Problem

• Automated breast density classification using deep learning
• Addressing class imbalance with novel loss function
• Enhancing mammography interpretation accuracy and consistency

Innovation

• Attention-enhanced CNNs for density classification
• Combined Focal Label Smoothing Loss function
• Optimized ensemble voting for superior performance