Using two palpable measurements improves the subject-specific femoral modeling

Area of Science:

• Biomedical Engineering
• Orthopedics
• Medical Imaging

Background:

• Subject-specific musculoskeletal models are crucial for biomedical research and clinical applications.
• Generating patient-specific models from CT/MRI is complex and requires specialized skills.
• Current methods often approximate bone models by scaling generic templates, limiting accuracy.

Purpose of the Study:

• To identify clinically available parameters (palpable measures, demographics) for improved femoral dimension prediction.
• To compare a novel non-homogenous scaling method against uniform scaling.
• To facilitate rapid and accurate generation of subject-specific femoral models.

Main Methods:

• Developed a non-homogenous anthropometric scaling method for femoral dimensions.
• Utilized clinically obtainable parameters, excluding imaging or invasive procedures.
• Compared predictive accuracy against homogenous (uniform) scaling.

Main Results:

• The non-homogenous scaling method significantly improved prediction of five key femoral measures compared to uniform scaling.
• Coefficient of determination (r²) for femoral depth increased from 0.22 (homogenous) to 0.60 (non-homogenous).
• Demonstrated enhanced accuracy in predicting femoral dimensions using non-invasive clinical data.

Conclusions:

• A non-homogenous scaling approach using clinical parameters offers a more accurate method for predicting femoral dimensions.
• This method enables the rapid and accurate creation of subject-specific musculoskeletal models without imaging.
• Clinical applicability for joint replacement, surgical planning, and gait analysis is enhanced.