Real-Time, Population-Based Reconstruction of 3D Bone Models via Very-Low-Dose Protocols

2025-08-19

AI Summary

Conventional CT-dependent skeletal modeling suffers from high radiation exposure, time-consuming manual segmentation, and limited adaptability—hindering preoperative planning and intraoperative guidance. To address these limitations, this paper proposes a real-time 3D bone model reconstruction method leveraging dual-plane X-ray imaging, eliminating reliance on CT scans and manual annotation. Our core contribution is SSR-KD, a semi-supervised knowledge distillation framework that synergistically integrates a small set of labeled X-ray images with a large volume of unlabeled data. The method achieves rapid reconstruction (≤30 seconds) with high geometric fidelity (mean surface-to-surface error <1.0 mm). Validation in high tibial osteotomy simulation demonstrates clinical usability comparable to CT-based ground truth. This work substantially reduces patient radiation dose and advances the clinical deployment of low-dose imaging for personalized orthopedic surgery.

Technology Category

• Computer Vision: 3D Computer Vision
• Machine Learning: Semi-Supervised Learning
• Planning, Routing, and Scheduling: Model-Based Reasoning

Application Category

• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

Patient-specific bone models are essential for designing surgical guides and preoperative planning, as they enable the visualization of intricate anatomical structures. However, traditional CT-based approaches for creating bone models are limited to preoperative use due to the low flexibility and high radiation exposure of CT and time-consuming manual delineation. Here, we introduce Semi-Supervised Reconstruction with Knowledge Distillation (SSR-KD), a fast and accurate AI framework to reconstruct high-quality bone models from biplanar X-rays in 30 seconds, with an average error under 1.0 mm, eliminating the dependence on CT and manual work. Additionally, high tibial osteotomy simulation was performed by experts on reconstructed bone models, demonstrating that bone models reconstructed from biplanar X-rays have comparable clinical applicability to those annotated from CT. Overall, our approach accelerates the process, reduces radiation exposure, enables intraoperative guidance, and significantly improves the practicality of bone models, offering transformative applications in orthopedics.

Problem

• Reconstructing 3D bone models from low-dose X-rays
• Reducing CT radiation exposure and manual delineation
• Enabling intraoperative bone modeling for surgical guidance

Innovation

• AI framework reconstructs bone models from X-rays
• Uses semi-supervised knowledge distillation technique
• Achieves sub-1mm accuracy in 30 seconds