Meta-heuristic optimization algorithms based feature selection for joint moment prediction of sit-to-stand movement...

Area of Science:

• Biomechanics
• Computational Science
• Clinical Research

Background:

• The sit-to-stand (STS) movement is crucial for daily function, requiring complex coordination of lower extremities and trunk.
• Accurate joint moment estimation is vital for biomechanical analysis but traditional methods are often restrictive or complex.
• Machine learning (ML) offers potential for joint moment estimation, yet efficient feature selection from diverse data remains a challenge.

Purpose of the Study:

• To develop a method for predicting joint moments during STS using minimal input data.
• To leverage metaheuristic optimization algorithms for effective feature selection in ML models.
• To enhance the accuracy of joint moment estimation for biomechanical and clinical applications.

Main Methods:

• Utilized motion analysis data from 20 participants with varying physical properties.
• Employed Manta Ray Foraging Optimization (MRFO), Marine Predators Algorithm (MPA), and Equilibrium Optimizer (EO) for feature selection.
• Applied Decision Tree Regression (DTR), Random Forest Regression (RFR), Extra Tree Regression (ETR), and eXtreme Gradient Boosting Regression (XGBoost Regression) for joint moment prediction.

Main Results:

• The Equilibrium Optimizer with Extra Tree Regression (EO-ETR) demonstrated superior performance for ankle, knee, and neck joint moment prediction.
• The Marine Predators Algorithm with Extra Tree Regression (MPA-ETR) showed the best results for hip joint moment prediction.
• The study successfully predicted joint moments using a minimal set of input features.

Conclusions:

• Metaheuristic optimization combined with ML regression models provides an effective approach for joint moment prediction during STS.
• This method offers a more efficient and less restrictive alternative to traditional techniques for joint moment analysis.
• The findings have significant implications for advancing biomechanical research and improving clinical assessments of movement.