Related Experiment Video
Updated: Jun 26, 2026
06:54
Photorealistic Learned Landscapes for Augmented Reality
Published on: June 27, 2025
Surface reconstruction from planar x-ray images using moving least squares.
T Cresson1, B Godbout, D Branchaud
1Ecole de Technologie Supèrieure, 1100 rue Notre-Dame Ouest, Montréal, Quebec, Canada.
Summary
This study introduces a novel method using moving least squares to create accurate 3D bone models from X-rays. This technique offers precise skeletal shape and posture measurements, comparable to CT scans.
Area of Science:
- Medical imaging
- Biomechanical engineering
- Computational anatomy
Background:
- Planar radiographs are the standard for assessing skeletal shape and posture.
- Accurate 3D bone modeling is crucial for clinical diagnosis and treatment planning.
Purpose of the Study:
- To develop a method for generating patient-specific 3D bone models from planar X-ray images.
- To evaluate the accuracy of the proposed 3D reconstruction technique.
Main Methods:
- An as-rigid-as-possible deformation approach utilizing moving least squares.
- Reconstruction of 3D bone models from biplane X-ray images.
- Interactive shape editing capabilities of the prototype.
Main Results:
- Successful generation of 3D personalized bone models from planar X-ray data.
- Demonstrated good accuracy in biplane reconstructions of the femur and vertebrae.
- Achieved accuracy comparable to computed tomography (CT) scans.
Conclusions:
- The proposed method provides an accurate and efficient way to obtain 3D bone models from standard X-rays.
- This approach has the potential to enhance the assessment of skeletal weight-bearing shape and posture.
- Interactive editing capabilities offer clinical utility for personalized bone modeling.
