Hybrid Deep Learning and Handcrafted Feature Fusion for Mammographic Breast Cancer Classification

2025-07-26

AI Summary

Distinguishing benign from malignant lesions in mammographic images remains challenging due to subtle visual differences, hindering accurate automated classification. To address this, we propose a lightweight hybrid feature fusion framework that jointly leverages deep features from ResNet-50, self-supervised representations from DINOv2, and handcrafted features (LBP and HOG), integrated via a Transformer-based multimodal embedding fusion module. Our key contribution is the empirical validation that handcrafted features effectively compensate for the fine-grained discriminative limitations of purely deep learning models—without increasing architectural complexity or parameter count. Evaluated on the CBIS-DDSM dataset, our method achieves an AUC of 79.6%, recall of 80.5%, and F1-score of 67.4%, approaching state-of-the-art performance while significantly improving computational efficiency.

Technology Category

Computer Vision: Multi-modal VisionMachine Learning: Multi-instance/Multi-view LearningIntelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphsSearch and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for rankingUser Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems

Abstract

Automated breast cancer classification from mammography remains a significant challenge due to subtle distinctions between benign and malignant tissue. In this work, we present a hybrid framework combining deep convolutional features from a ResNet-50 backbone with handcrafted descriptors and transformer-based embeddings. Using the CBIS-DDSM dataset, we benchmark our ResNet-50 baseline (AUC: 78.1%) and demonstrate that fusing handcrafted features with deep ResNet-50 and DINOv2 features improves AUC to 79.6% (setup d1), with a peak recall of 80.5% (setup d1) and highest F1 score of 67.4% (setup d1). Our experiments show that handcrafted features not only complement deep representations but also enhance performance beyond transformer-based embeddings. This hybrid fusion approach achieves results comparable to state-of-the-art methods while maintaining architectural simplicity and computational efficiency, making it a practical and effective solution for clinical decision support.

Problem

• Improving mammographic breast cancer classification accuracy
• Combining deep and handcrafted features for better performance
• Enhancing clinical decision support with hybrid fusion approach

Innovation

• Hybrid deep learning and handcrafted feature fusion
• ResNet-50 backbone with transformer embeddings
• Enhanced performance with computational efficiency