Decoding Upper-Limb Movement Intention Through Adaptive Dynamic Movement Primitives: A Proof-of-Concept Study with a...

Area of Science:

• Robotics
• Biomechanics
• Rehabilitation Engineering

Background:

• Upper-limb impairments significantly affect daily activities.
• Exoskeleton assistance can aid individuals with residual movement capabilities.
• Accurate intention decoding is crucial for effective human-robot collaboration.

Purpose of the Study:

• To develop and validate an intention decoding algorithm for controlling a shoulder-elbow exoskeleton.
• To assist users with upper-limb impairments in completing reaching tasks.
• To predict movement phase and hand goal position in real-time.

Main Methods:

• Utilized adaptive Dynamic Movement Primitives and Gaussian Mixture Models.
• Algorithm processed initial joint angle measures for movement analysis.
• Tested on a healthy subject using a 4 degrees-of-freedom shoulder-elbow exoskeleton in planar reaching tasks.

Main Results:

• Predicted hand's final position with <5 cm error after 0.25s.
• Achieved final positions within 4 cm of target on average.
• Demonstrated reduced Biceps Brachii activation and faster task completion times.

Conclusions:

• The developed algorithm effectively decodes user intention for exoskeleton-assisted reaching.
• This technology shows promise for improving functional independence in individuals with upper-limb impairments.
• Real-time intention decoding enhances human-robot interaction and task efficiency.