Real-Time Musculoskeletal Kinematics and Dynamics Analysis Using Marker- and IMU-Based Solutions in Rehabilitation

Area of Science:

• Biomechanics
• Musculoskeletal Modeling
• Real-time Motion Analysis

Background:

• Accurate estimation of kinematic and dynamic quantities is crucial for understanding human movement.
• Existing methods often require controlled laboratory environments and offline processing.
• Bridging the gap between laboratory accuracy and real-world applicability is a significant challenge.

Purpose of the Study:

• To develop and validate a real-time framework for estimating musculoskeletal quantities.
• To enable estimations using both marker-based and inertial measurement units (IMUs).
• To reduce computational latency while maintaining accuracy compared to offline methods.

Main Methods:

• Subject-specific musculoskeletal models were employed.
• Calculations included kinematics, generalized forces, muscle forces, joint reaction loads, and ground reaction wrenches during walking.
• Emphasis was placed on filtering and differentiation techniques for noisy data.
• The framework was designed for marker-based and IMU data integration.

Main Results:

• Real-time estimates of joint moments, muscle forces, and reaction loads closely matched offline OpenSim analyses.
• Model-based ground reaction wrench estimation highlighted sensitivity to errors.
• The system demonstrated successful application in rehabilitation and gait retraining contexts.
• Computational latency was significantly reduced while preserving accuracy.

Conclusions:

• The developed open-source framework enables accurate real-time estimation of musculoskeletal dynamics.
• The system extends motion analysis beyond laboratory settings, applicable in diverse conditions.
• This framework represents a significant advancement for tackling complex biomechanical challenges.
• Future applications include predicting environmental interactions for enhanced rehabilitation and gait analysis.