SPIDER: Structure-Preferential Implicit Deep Network for Biplanar X-ray Reconstruction

2025-07-07

AI Summary

Dual-plane X-ray reconstruction of 3D CT volumes suffers from incomplete bone structures, blurred tissue boundaries, and anatomical distortions. Method: We propose an anatomy-aware implicit neural representation method that integrates hierarchical anatomical priors into a unified encoder-decoder architecture, enabling pixel-wise joint modeling of X-ray intensity and anatomical semantics within the implicit decoder. The approach employs a supervised deep learning framework using only two orthogonal X-ray projections as input. Contribution/Results: Evaluated on a clinical cranial dataset, our method reconstructs high-fidelity, anatomically accurate CT volumes from just two projections. Experiments demonstrate substantial improvements in bone continuity and tissue boundary precision; downstream segmentation achieves an average 8.2% Dice score improvement over baselines. The method shows strong potential for clinical applications including surgical planning and postoperative assessment.

Technology Category

• Computer Vision: Representation Learning for Vision
• Machine Learning: Representation Learning
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Biplanar X-ray imaging is widely used in health screening, postoperative rehabilitation evaluation of orthopedic diseases, and injury surgery due to its rapid acquisition, low radiation dose, and straightforward setup. However, 3D volume reconstruction from only two orthogonal projections represents a profoundly ill-posed inverse problem, owing to the intrinsic lack of depth information and irreducible ambiguities in soft-tissue visualization. Some existing methods can reconstruct skeletal structures and Computed Tomography (CT) volumes, they often yield incomplete bone geometry, imprecise tissue boundaries, and a lack of anatomical realism, thereby limiting their clinical utility in scenarios such as surgical planning and postoperative assessment. In this study, we introduce SPIDER, a novel supervised framework designed to reconstruct CT volumes from biplanar X-ray images. SPIDER incorporates tissue structure as prior (e.g., anatomical segmentation) into an implicit neural representation decoder in the form of joint supervision through a unified encoder-decoder architecture. This design enables the model to jointly learn image intensities and anatomical structures in a pixel-aligned fashion. To address the challenges posed by sparse input and structural ambiguity, SPIDER directly embeds anatomical constraints into the reconstruction process, thereby enhancing structural continuity and reducing soft-tissue artifacts. We conduct comprehensive experiments on clinical head CT datasets and show that SPIDER generates anatomically accurate reconstructions from only two projections. Furthermore, our approach demonstrates strong potential in downstream segmentation tasks, underscoring its utility in personalized treatment planning and image-guided surgical navigation.

Problem

• Reconstructs 3D CT volumes from two X-ray projections
• Improves bone geometry and tissue boundary accuracy
• Reduces soft-tissue artifacts using anatomical constraints

Innovation

• Implicit neural representation with anatomical priors
• Joint supervision for intensity and structure learning
• Embedding anatomical constraints to reduce artifacts