Back to search

Georgia Institute of Technology

Simulation of agonist and antagonist muscle activation patterns in bidirectional postural perturbation in cats

Abstract

dc:description.abstract

We studied the effects of varying perturbation magnitude and direction on the postural control process of the central nervous system (CNS) caused by perturbation, before and after sensory loss. The electromyogram (EMG) response to a postural perturbation can be composed by a weighted sum of the center of mass (CoM) kinematics. We extended an existing CoM feedback model which predicted EMG of one muscle for unidirectional perturbations; we used recorded data of bidirectional perturbations, which caused muscle activity in anterior as well as posterior muscles. Modeling the CNS as two delayed feedback controllers, we reconstructed the EMGs of two antagonistic muscles simultaneously that were recorded during postural perturbation experiments on cats. Minimizing the error between predicted and recorded EMG and CoM kinematics, we were able to identify controller gains that would result in the best prediction of the recorded EMGs. We hypothesized that the weights on the CoM kinematics remained constant independent of variations in perturbation magnitude or reversed perturbation direction. We applied our model to data from bidirectional perturbations with varying magnitude, with which the cats were perturbed for a short time in one direction and a longer time in the opposite direction. The gains showed small variation for EMG predictions following long perturbations; however, the prediction of EMG following the initial displacement resulted in large gain variations. We showed that these variations were caused by our optimization methods, which was not able to consistently identify controller gains for short initial movements. Using the weights identified for unidirectional perturbations, we were able to predict muscle activity for both directions with the same gains. This suggests that the weights of the CoM kinematics for each muscle did not change for varying perturbation magnitude. We conclude that varying EMG shapes were induced solely by the variation of the CoM kinematics. We repeated the investigations on data that was recorded from cats suffering from sensory loss and found reduced CoM acceleration feedback.

Degree

thesis:*
Department dc:contributor.department
Electrical and Computer Engineering
Grantor dc:publisher
Georgia Institute of Technology
Year dc:date.issued
2006

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Koenig, Alexander C.
Advisor dc:contributor.advisor
  • Ting, Lena H.
Committee members dc:contributor.committeemember
  • Egerstedt, Magnus
  • Lee, Robert

Subjects

dc:subject × 6

Rights

Language dc:language.iso
en_US

Identifiers

dc:identifier.*
Handle dc:identifier.uri
http://hdl.handle.net/1853/11576

Chain of custody

source
Harvested from
Georgia Tech
Base URL
repository.gatech.edu/server/oai/request
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

Koenig, Alexander C.. Simulation of agonist and antagonist muscle activation patterns in bidirectional postural perturbation in cats. Georgia Institute of Technology, 2006. http://hdl.handle.net/1853/11576