Myoelectric Control and Neural Decoding

Efforts to develop brain machine interfaces, be they invasive or noninvasive, are often challenged by the need to maximize decoding accuracy while minimizing computational complexity. Our lab develops real-time neural decoding strategies that can be used to estimate kinematic properties of human movement. Our most recent efforts have centered on the use of residual muscle activity in the amputated limb to continuously predict the control of lower limb prostheses to provide natural movement across tasks (walking, stairs, etc.).

 

  • Neural Prosthetic
  • Weight-Sensored Stair
  • In-Software Controls
  • Output Data

Nerual Prosthetic used in research

Close-up of weight-sensored stair used in research

Screenshot of Neural Network Training experimental controls

Neural Network Training software data display

Selected Publications 

 

*Zabre-Gonzalez E., *Riem L., Voglewede P., Silver-Thorn B., Koehler-McNicholas S., Beardsley S. A., (2021), Continuous myoelectric prediction of future ankle angle across ambulation conditions and their transitions, Front. Neurosci. - Neuroprosthetics., 15, p. 1-13., doi: 10.3389/fnins.2021.709422. PMID: 34483828; PMCID: PMC8416349.

*Zabre-Gonzalez E. V., *Amieva-Alvarado D., Beardsley S. A., (2021), Prediction of EMG activation profiles from gait kinematics and kinetics during multiple terrains, Annu Int Conf IEEE Eng Med Biol Soc. 2021 Nov;2021:6326-6329. doi: 10.1109/EMBC46164.2021.9630067. PMID: 34892560.

Farmer, S., Silver-Thorn B., Voglewede P., Beardsley S.A. (2014), "Within-socket myoelectric prediction of continuous ankle kinematics for control of a powered transtibial prosthesis," J. Neural Eng., 11(5):056027. DOI: 10.1088/1741-2560/11/5/056027

 

 

View more INSL Research