Static torque-angle relation of human elbow joint estimated with artificial neural network technique

Area of Science:

• Biomechanics
• Neuroscience
• Musculoskeletal System

Background:

• Understanding the relationship between joint angles, muscle activity, and joint torque is crucial for biomechanical analysis.
• Previous models often simplify the complex, non-linear dynamics of human joints.

Purpose of the Study:

• To investigate the static relationship between elbow joint angle and torque under constant muscle activity.
• To model this relationship using artificial neural networks (ANNs) and analyze the torque-angle behavior.

Main Methods:

• Utilized an artificial neural network technique with integrated electromyograms (IEMGs) and joint angles as inputs, and elbow joint torque as the output.
• Subjects maintained specific elbow joint angles while forearm movement was influenced by gravitational force.
• Employed backpropagation learning for the ANN model to estimate torque-angle relations.

Main Results:

• The elbow joint torque demonstrated a non-linear pattern, increasing and then decreasing with joint extension.
• The torque-angle relationship deviates from simple spring-like behavior when the forearm is displaced from an equilibrium point.
• The moment arm of elbow flexor muscles appears to significantly influence the observed torque-angle dynamics.

Conclusions:

• The elbow joint's torque-angle relationship is complex and not analogous to a simple spring.
• Musculoskeletal structure, particularly the moment arm of flexor muscles, plays a dominant role in determining joint torque.
• ANNs provide a powerful tool for modeling these intricate biomechanical relationships.