Elbow joint angle and elbow movement velocity estimation using NARX-multiple layer perceptron neural network model...

Area of Science:

• Biomechanics
• Neuroscience
• Biomedical Engineering

Background:

• Elbow dynamics estimation is a key area in biomechanical research.
• Accurate measurement of elbow kinematics is crucial for various applications.

Purpose of the Study:

• To propose a solution for estimating elbow movement velocity and joint angle from Surface Electromyography (SEMG) signals.
• To develop a model for predicting elbow kinematics using muscle activity.

Main Methods:

• SEMG signals were acquired from the biceps brachii muscle.
• Integrated EMG (IEMG) and Zero Crossing (ZC) parameters were extracted.
• A Nonlinear Auto Regressive with eXogenous inputs (NARX) based Multiple Layer Perceptron Neural Network (MLPNN) model was developed and trained.

Main Results:

• The proposed NARX MLPNN model demonstrated accurate estimation of elbow joint angle and movement angular velocity.
• Validation using regression coefficient (R) showed high accuracy: 0.9641 for angular displacement and 0.9347 for angular velocity.

Conclusions:

• The NARX MLPNN model is effective for estimating elbow angular displacement and movement angular velocity.
• This approach offers a reliable method for non-invasive elbow kinematics assessment.