Related Experiment Video
Updated: Jun 18, 2026
11:16
Engineering Platform and Experimental Protocol for Design and Evaluation of a Neurally-controlled Powered Transfemoral Prosthesis
Published on: July 22, 2014
Development of a myoelectric control scheme based on a time delayed neural network
Alan Smith1, Pooja Nanda, Edward E Brown
1Department of Electrical Engineering, Rochester Institute of Technology, Rochester, NY, USA.
Summary
This study explores a myoelectric control system using a time-delayed neural network (TDNN) for robotic rehabilitation. It demonstrates effective control of multiple joint movements using electromyographic (EMG) signals.
Area of Science:
- Biomedical Engineering
- Robotics
- Neuroscience
Background:
- Rehabilitation robotics require intuitive control schemes.
- Myoelectric control using electromyographic (EMG) signals offers a promising avenue.
- Time-delayed neural networks (TDNNs) have potential for processing complex biological signals.
Purpose of the Study:
- To investigate a novel myoelectric control scheme for rehabilitation robotics.
- To evaluate the efficacy of a TDNN in translating EMG signals into joint position commands.
- To assess the feasibility of controlling multiple degrees of freedom simultaneously.
Main Methods:
- Utilized four electromyographic (EMG) signals from elbow and shoulder muscles as input.
- Employed a time-delayed neural network (TDNN) for signal processing.
- The TDNN output predicted elbow and shoulder joint positions in the sagittal plane.
Main Results:
- Demonstrated the ability to control multiple degrees of freedom concurrently.
- Achieved comparable position prediction accuracy with a 300ms delay and 100ms interval, differing from prior optimal settings.
- Indicated the feasibility of a TDNN-based control scheme for rehabilitation applications.
Conclusions:
- A TDNN-based myoelectric control scheme is feasible for rehabilitation robotics.
- Reduced time delays in TDNNs can yield effective multi-joint control.
- This approach holds potential for enhancing robotic-assisted rehabilitation therapies.
