Thakor Lab YouTube page
 
 
 
Brain-Computer Interfaces

& Neuroprostheses
EMG Decoding
 Abstract
Despite advances in invasive brain computer interfaces, the current staple for prosthetic control is to use electromyography (EMG). Myoelectric signals are acquired through skin surface electrodes, and involve the detection of electrical activity that corresponds to muscle contraction. Through various training protocols and signal processing techniques, limited EMG control of powered prosthetic devices is possible. Our work focuses on developing more functional control schemes as well as more intuitive interfaces for users of these powered upper-limb prostheses.
 
EMG-controlled gaming. The left screen shows sample EMG processing.
 
In the past, powered upper-limb prostheses have been limited mechanically by devices with very few degrees of freedom. However, the ongoing development of multi-fingered dexterous hands, and improvement in electrode technology, has precipitated the need for improved control algorithms for next generation prostheses. Using both able-bodied and amputee subjects, our research focuses on high-accuracy decoding of a wide range of discrete finger and grasp movements to improve the range of functionality.
 

By using the Cyberglove, it becomes possible to continuously track hand conformation such as the joint angle of the metacarpophalangeal joints (MCP) shown here.
 
Currently we are pursuing long-term improvement of the human-prosthesis relationship. One approach we are taking is the development of smarter prosthetic devices incorporating radio-frequency identification for improved interaction with the environment. Additionally, we are investigating adaptive pattern recognition strategies that allow the prosthesis to learn its user's intentions and tendencies over time, reducing the burden on the user for supervised device training.

We are also seeking to expand the possibilities of Human-Computer Interaction through myoelectric control. Development of myoelectric cursor control in combination with a smarter prosthesis may allow amputees to one day interact with their PC without ever touching a mouse.

Through our collaborations with Infinite Biomedical Technologies, we are able to implement our ideas developed through research into functional prosthetic limbs that are frequently tested and worn by amputees.

Current projects are: (1) Novel pattern recognition techniques, (2) Incorporation of RFID into prostheses, (3) Tactile Sensing, and (4) Development of a virtual training environment
 
Researchers
Ryan Smith
Luke Osborn
Matt Masters
Joseph Betthauser
Bobby Beaulieu
 
Collaborators
Infinite Biomedical Technologies
Charles Dankmeyer - Dankmeyer Prosthetics
 
Funding
Defense Advanced Project and Research Agency (DARPA) - contract N66001-06-C-8005
 
Publications
Tenore F, Ramos A, Acharya S, Etienne-Cummings R and Thakor NV, Decoding of individuated finger movements using surface electromyography, IEEE Trans Biomed Eng, in press, 2008

Smith RJ, Tenore F, Huberdeau D, Thakor NV, Continuous Decoding of Finger Position from Surface EMG Signals for Control of Powered Prostheses, Conf Proc IEEE Eng Med Biol Soc, 1:197-200, 2008

Huberdeau D, Aggarwal V, Tenore F, Fritz K, Etienne-Cummings R, Thakor NV, Real-time finger tracking to improve upper-limb prosthetics control, Conf Proc Northeast Bioeng Conf, 2008

Tenore F, Ramos A, Fahmy A, Acharya S, Etienne-Cummings R, Thakor NV, Towards the Control of Individual Fingers of a Prosthetic Hand Using Surface EMG Signals, Conf Proc IEEE Eng Med Biol Soc, 1:6145-8, 2007
 
Go to Top
 
 
   

Copyright © 2008 JHU Neuroengineering & Biomedical Instrumentation Lab
All rights reserved.