Shape model constrained scaling improves repeatability of gait data.

Area of Science:

• Biomechanics
• Human Movement Analysis
• Computational Modeling

Background:

• Variability in gait research often stems from researcher decisions in creating kinematic and musculoskeletal models.
• Statistical shape models offer accurate bone geometry prediction, potentially improving clinical gait analysis repeatability.

Purpose of the Study:

• To evaluate if shape model scaling improves kinematic and kinetic gait data repeatability compared to traditional linear scaling methods.
• To quantify the impact of different scaling techniques on gait data variability.

Main Methods:

• Five participants underwent motion capture, including static and walking trials.
• Five researchers generated kinematic models using linear scaling (OpenSim) and shape-model scaling (MAP Client).
• Inverse kinematic and dynamic analyses were performed, with inter-researcher variation assessed.

Main Results:

• Linear scaling resulted in significantly higher variability in joint angles, moments, and powers (P < 0.001) compared to shape-model scaling.
• Variation was at least three times greater with linear scaling than with shape-model scaling.
• Shape-model scaling demonstrated superior repeatability for gait data.

Conclusions:

• Linear scaling introduces substantial variability in gait data across researchers, even with identical experimental data.
• Employing statistical shape models for scaling musculoskeletal models leads to more repeatable kinematic and kinetic gait data.
• Shape-model scaling is recommended for improving consistency in clinical gait analysis.