Semi-automated stereoradiographic upper limb 3D reconstructions using a combined parametric and statistical model: a...

Area of Science:

• Orthopedic surgery
• Medical imaging
• Biomechanical engineering

Background:

• Quantitative assessment of 3D clinical indices is vital for elbow surgery planning.
• 3D parametric modeling using bi-planar radiographs has been successful for spine and lower limb analysis.
• This technique offers a potential alternative to CT scans, reducing radiation exposure.

Purpose of the Study:

• To adapt and validate a 3D parametric modeling method for upper limb (humerus) analysis using bi-planar radiographs.
• To assess the accuracy and reproducibility of this method for clinical applications.
• To provide a low-radiation alternative for pre-operative evaluation in elbow arthroplasty.

Main Methods:

• 3D parametric models were created from CT scans of 20 cadaveric humerus, radius, and ulna specimens.
• Correlations between geometric primitive descriptors were established.
• A semi-automated reconstruction method using bi-planar radiographs was developed to generate personalized 3D humerus models.

Main Results:

• The 3D reconstruction achieved a mean points-to-surface distance of 0.9 mm (RMS = 2.5 mm).
• Computed clinical parameters showed mean differences of 0.4–1.9 mm and 1.7°–2.3° compared to CT scans.
• The reconstruction process took approximately 5 minutes, demonstrating efficiency.

Conclusions:

• The developed method accurately quantifies upper limb morphological parameters with minimal radiation exposure.
• Preliminary validation indicates the technique is sufficiently fast and accurate for routine clinical use.
• This approach serves as a viable alternative to CT scans for pre-operative planning in total elbow arthroplasty.