Systems, Man, and Cybernetics

 

Home

• [P960] Development of a Full Body Balance Model Using an Artificial Neural Network Approach
  Author(s): Roseann Trevino, Michael T. Frye, Chunjiang Qian
  
  

  Abstract
  The purpose of this paper is to identify body balance using an artificial neural network approach. This research entails the study of dynamic stability within a normal person. This study is inspired because persons suffering from lower extremity loss suffer a variety of complications including numbness on the residual limb and sores caused from the prosthetic. Because this occurs, they have a slightly abnormal gait pattern, possibly to keep balance while in motion. This study analyzes the gait motion of a normal healthy subject. We take the data and manipulate it to delete or alter the function of the right leg. Data was taken using an 8 camera VICON motion capture system at the Andrew Gitter GAIT Laboratory located in the Audie L. Murphy Veterans hospital. The markers placed at joints of the body were captured to give a 3-D position at a sampling rate of 120 MHz. A Neural Network was used for the modeling of normal walking gait using the given data.

  
   

Full Text (PDF)