Investigation of trunk muscle activities during lifting using a multi-objective optimization-based model and...

Area of Science:

• Biomechanics
• Musculoskeletal Modeling
• Spinal Physiology

Background:

• Accurate prediction of trunk muscle activity is crucial for understanding spinal loading during lifting.
• Existing models often lack comprehensive consideration of spinal stability and in vivo physiological data.

Purpose of the Study:

• To develop and validate a six-degree-of-freedom musculoskeletal model of the lumbar spine.
• To predict trunk muscle activity during various lifting intensities (light, moderate, heavy) in a standing posture.
• To compare the efficacy of Vector Evaluated Particle Swarm Optimization (VEPSO) and Nondominated Sorting Genetic Algorithm (NSGA) in this predictive modeling.

Main Methods:

• Formulation of a multi-objective optimization problem balancing muscle stress minimization and spinal stability maximization.
• Application of VEPSO and NSGA algorithms to solve the optimization problem.
• Validation of model predictions against in vivo intradiscal pressure and electromyography (EMG) data.

Main Results:

• Both VEPSO and NSGA predicted antagonistic abdominal muscle co-activity and increased spinal stability with lifting load.
• Muscle activity predictions from both algorithms were consistent and correlated well with in vivo EMG data across all tasks.
• The model successfully reproduced in vivo intradiscal pressure, indicating physiological relevance.

Conclusions:

• The developed six-degree-of-freedom musculoskeletal model provides improved estimations of trunk muscle activity during lifting.
• Incorporating spinal stability and in vivo intradiscal pressure enhances the predictive accuracy of musculoskeletal models.
• VEPSO and NSGA are effective algorithms for solving complex biomechanical optimization problems in spinal modeling.