Clinical Contrast-Enhanced Computed Tomography With Semi-Automatic Segmentation Provides Feasible Input for...

Area of Science:

• Biomechanics
• Medical Imaging
• Computational Modeling

Background:

• Osteoarthritis (OA) progression is linked to joint function and tissue failure.
• Current joint geometry segmentation for computational models is manual, time-consuming, and not clinically practical.

Purpose of the Study:

• To evaluate a semi-automatic segmentation method for tibial and femoral cartilage as input for finite element (FE) models.
• To assess the clinical applicability of this novel segmentation technique.

Main Methods:

• Knee joints from seven volunteers were imaged using contrast-enhanced computed tomography (CT).
• Cartilage segmentation was performed using both semi-automatic and manual methods.
• Fibril-reinforced poroviscoelastic (FRPVE) FE models were generated and mechanically analyzed under physiological loading.

Main Results:

• The semi-automatic method accelerated segmentation by over 90%.
• Negligible differences were observed in maximum principal stress (<1 MPa), strain (<0.72%), and fibril strain (<0.40%) between semi-automatic and manual models.
• No statistically significant differences (p >0.05) were found in contact areas, forces, pore pressures, or average strains.

Conclusions:

• The CT-based semi-automatic segmentation method is a viable and efficient tool for FE modeling.
• This method offers a faster alternative to manual segmentation for clinical evaluation of knee joint function in OA.