top of page

Milestones

Ground Truth Project 

Screen Shot 2020-04-27 at 12.46.10 PM.pn
Screen Shot 2020-04-27 at 12.55.06 PM.pn

  • Using the gait events found from the GRF, stance time asymmetry during walking trials was calculated and normalized 

  • Acceleration data was segmented into gait cycles at the heel strike locations using the gait events from the force data

  • These two sets of data were used to train an asymmetry prediction model.   

Acceleration Project 

Screen Shot 2020-04-27 at 12.46.51 PM.pn

  • The importance of this project was to determine the feasibility of segmenting acceleration gait cycles without the use of GRF for future incorporation into a model-based algorithm used to detect asymmetric gait outside of a laboratory setting

  • To determine placement of segmentation, GRF were used to visually determine where gait events appear in the acceleration data for all directions (x, y, z, and the magnitude)

  • The z direction (upward/downward) was chosen as the best option for determining gait events due to heel strike appearing at a recognizable maximum prior to a prevalent minimum 

  • The frame location of heel strike was first defined in the z direction acceleration (left figure) then displayed on the magnitude (right figure) 

  • The defined heel strike locations in the magnitude were used to split the acceleration signals into gait cycles

bottom of page