Using additive manufacturing in accuracy evaluation of reconstructions from computed tomography.

Area of Science:

• Orthopedics
• Biomedical Engineering
• Medical Imaging

Background:

• Patient-specific bone models fabricated via additive manufacturing offer significant potential for surgical planning, medical training, and research.
• Computed tomography (CT) is a common imaging modality for creating digital bone models.

Purpose of the Study:

• To evaluate the accuracy and reproducibility of 3D bone surface reconstruction for hip and shoulder joints using CT data.
• To identify critical factors influencing the accuracy of patient-specific bone models.

Main Methods:

• 3D virtual models were generated from CT image segmentation of cadaveric hip and shoulder joints.
• Fused deposition modeling was used to fabricate physical bone models.
• Laser scanning was employed to compare the accuracy of segmented models, physical models, and original cadaver bones.

Main Results:

• The overall bone reconstruction process demonstrated a reproducibility of 0.3 ± 0.4 mm.
• Physical model production achieved high accuracy (0.1 ± 0.1 mm), while segmentation accuracy was 0.3 ± 0.4 mm.
• Segmentation accuracy was identified as the primary determinant of overall reconstruction fidelity. Articular surface shapes were generally well-reproduced, with decreased accuracy at the periphery, especially near periosteum and osteophytes.

Conclusions:

• Accurate 3D bone reconstruction from CT scans is feasible for hip and shoulder joints.
• Image segmentation quality is the critical factor limiting the accuracy of patient-specific additive manufactured bone models.
• The study highlights the potential of these models for clinical applications while noting limitations in areas with complex anatomical features.