Invisible Yet Detected: PelFANet with Attention-Guided Anatomical Fusion for Pelvic Fracture Diagnosis

2025-09-17

AI Summary

Pelvic fractures often appear occult on conventional X-ray radiographs, leading to high misdiagnosis rates. To address this, we propose PelFANet, a dual-stream attention network that jointly processes raw X-ray images and semantic segmentation–guided bone mask images. PelFANet introduces Fusion Attention Blocks (FABlocks) to enable iterative cross-stream feature interaction and enhancement, and employs an anatomy-aware two-stage training strategy—enabling strong generalization to *unseen* fracture types without requiring any “invisible fracture” samples during training. The framework performs end-to-end fracture classification without manual feature engineering. Evaluated on the AMERI dataset, PelFANet achieves 88.68% accuracy (AUC = 0.9334) for visible fractures and 82.29% accuracy (AUC = 0.8688) for previously unseen invisible fractures—substantially outperforming existing methods.

Technology Category

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

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Pelvic fractures pose significant diagnostic challenges, particularly in cases where fracture signs are subtle or invisible on standard radiographs. To address this, we introduce PelFANet, a dual-stream attention network that fuses raw pelvic X-rays with segmented bone images to improve fracture classification. The network em-ploys Fused Attention Blocks (FABlocks) to iteratively exchange and refine fea-tures from both inputs, capturing global context and localized anatomical detail. Trained in a two-stage pipeline with a segmentation-guided approach, PelFANet demonstrates superior performance over conventional methods. On the AMERI dataset, it achieves 88.68% accuracy and 0.9334 AUC on visible fractures, while generalizing effectively to invisible fracture cases with 82.29% accuracy and 0.8688 AUC, despite not being trained on them. These results highlight the clini-cal potential of anatomy-aware dual-input architectures for robust fracture detec-tion, especially in scenarios with subtle radiographic presentations.

Problem

• Diagnosing subtle pelvic fractures invisible on radiographs
• Fusing raw X-rays with segmented bone images
• Improving classification accuracy for challenging fracture cases

Innovation

• Dual-stream attention network fuses X-rays
• Fused Attention Blocks refine anatomical features
• Two-stage segmentation-guided training improves classification