Figure 2: COM trajectories of Subject 4 during stead-state gait.

1
P8
Validation of Whole Body COM Calculation in
Children Using the Gait Initiation Task
James Carollo, Ph.D., P.E., Tim Nicklas, B.S.,
Frank Chang, M.D.
Center for Gait and Movement Analysis (CGMA) The
Childrens Hospital / University of Colorado HSC
College of Engineering University of Colorado,
Boulder
Introduction
Methodology 3D Kinematic COM Calculation
Discussion

• A plot (using Igor Pro) of the whole body COM
  projected into the ground plane in quiet standing
  shows a displaced initial COM relative to the
  calculated COP for both the Vicon PlugIn and
  Johan model. Data for Subject 4 shown (12yr.
  female)

• The whole body center of mass (COM) represents
  the net result of all instantaneous limb
  movements and force production during a specific
  task.
• The whole body COM can be used to assess both
  dynamic balance and overall gait performance.
• During quiet standing, the projection of the COM
  in the ground plane coincides with the center of
  pressure (COP) - this provides an excellent
  empirical method for validating COM models.

• The correction to the Johan model adjusts the
  percent contribution of the segment masses so the
  total sums to 100.
• Once the correction was applied, the whole body
  COM was found to have a lower mean error at gait
  initiation than the other 2 models for the 6
  subjects (Table 2).
• The effect of this correction on steady state
  walking can be seen as the COM position no longer
  migrates out of the volume of the pelvis as the
  subject walks away from the coordinate system
  origin (Figure 2).

• A 13-segment whole body kinematic model was
  processed.
• The whole body COM was then calculated using the
  Johan model by Eames et al and the Vicon PlugIn
  default model.
• The COMs of each body segment were summed to find
  the whole body COM.

Table 1 Percent contribution of segments to
whole body COM in Johan and corrected Johan
models.
Clinical Significance
Methodology Kinetic COP Calculation

• The use of gait initiation as an appropriate gait
  task to characterize accuracy of COM calculation
  is demonstrated.
• The gait initiation task was used to identify a
  slight inaccuracy in the calculated whole body
  COM for subjects under 14 years of age in a
  commonly used full body model (Johan.mod).
• A correction factor was applied to this
  calculation and the improved COM trajectory is
  shown.

• A COP was calculated for the entire force
  platform array by applying moment balance
  equations to the total force and COP data for the
  individual force platforms.
• Individual force platform total force and COP
  data were exported from Vicon Workstation.

Methodology Functional Gait Tasks

• Subjects performed two functional gait tasks.
• Gait initiation subject remained in quiet
  standing with each foot on a separate force
  platform (2 and 3) in the 1-2-1 force platform
  array. When a visual cue was given, the patient
  initiated gait.
• Steady-state gait subject walked down a 30 ft
  walkway at a comfortable self-selected pace.
• The total force platform COP was compared to the
  floor plane projection of each whole body COM in
  quiet standing for COM validation in gait
  initiation.

Methodology Data Capture

• Data from 6 normal subjects participating in a
  larger study were selected.
• The data collection began with the subjects
  having their height, weight, and distance between
  several bony landmarks measured.
• 35 retro-reflective skin markers were placed on
  areas of the subjects head, chest, arms, and
  legs bilaterally.
• A 6-camera Vicon 512 kinematic measurement system
  was used to record and digitize 3-dimensional
  displacements of the markers.
• The ground reaction forces were recorded using 4
  Kistler Model 9281C piezoelectric force platforms
  embedded in the lab floor.

• A correction to the calculation of whole body COM
  for children under 14 years of age, with
  anthropometrics scaled to age, was applied to the
  Johan model.
• The total percentage at ages 0 to 13 was 97.7.
• Once this correction was applied, the whole body
  COM was found to have a lower mean error at gait
  initiation than the other two models, and the
  average relative position of the whole body COM
  to a fixed point on the pelvis became constant.
• The same correction was applied to the whole body
  COM for subjects 14 years of age and over, but
  the change was negligible.

Figure 2 COM trajectories of Subject 4 during
stead-state gait.
Results
Conclusion

• An error in the whole body COM for children under
  14 years of age with the uncorrected Johan model
  can be detected as a small shift in the
  calculated COM starting point during the quiet
  standing portion of gait initiation.
• The greater the distance the COM travels from the
  fixed coordinate system origin, the greater the
  error. This creates a problem when calculating
  whole body COM across large areas during steady
  state walking.
• These results support the inclusion of the task
  of gait initiation in any COM walking trial as a
  simple method to ensure system/model accuracy,
  and to validate the whole body COM.

Table 2 Distance between COM and COP in quiet
standing.
Acknowledgement
Christine Irelan, M.D. and Nancy Denniston,
M.S., M.A.
Figure1 COM trajectories of Subject 4 during
gait initiation.
8th Annual Meeting of the Gait and Clinical
Movement Analysis Society