Comparative analysis of methods for estimating arm segment parameters and joint torques from inverse dynamics

Area of Science:

• Biomechanics
• Human Movement Analysis
• Kinetics

Background:

• Accurate estimation of joint torques in upper limb movements is crucial for biomechanical analysis.
• Inaccuracies in kinematic data, body segment parameters (BSPs), and biomechanical models can affect joint torque calculations.
• Uncertainty in BSP estimation is particularly impactful for high-acceleration movements.

Purpose of the Study:

• To assess the sensitivity of inverse dynamics analysis to various sources of inaccuracy in planar arm movements.
• To quantify the influence of different BSP estimation methods on joint torque calculations.
• To compare the impact of BSP estimation methods on arm movement analysis versus gait analysis.

Main Methods:

• Performed a sensitivity analysis on planar arm reaching movements.
• Utilized eight regression models and one water immersion method for BSP estimation.
• Calculated normalized joint torque peaks and root mean square (rms) torques using a double-link arm model with a floating shoulder.
• Conducted statistical analyses to determine the influence of BSP models on kinetic variable variance.

Main Results:

• The method used for BSP estimation significantly influenced the normalized rms of joint torques.
• The interaction between BSP estimation, subject somatotype, and movement kinematics introduced significant variance.
• Variance in arm segment parameter estimation had a greater impact on calculated joint torques than kinematic variable variance.
• The influence of BSP estimation methods on arm movement inverse dynamics is greater than in gait analysis.

Conclusions:

• The choice of body segment parameter estimation method is a critical factor in the inverse dynamics analysis of upper limb movements.
• Normalized joint torque parameters are valuable for comparing the effects of different BSP estimation methods across diverse populations.
• Inverse dynamics of arm movements are more sensitive to BSP estimation than leg movements due to lower limb moments of inertia.