Figure 1.
A. Cartoon of experimental setup showing link-segment model, reflective marker locations, and position of subject relative to the projection screen mounted on a rigid wall. Circles indicate frontal plane view of 10-cm diameter fixed targets and 5-cm diameter moving target; B. Calibration trial had the hat placed on a high stool. H1, H2, and H3 are the reflective markers on the hat brim comprising the rigid body from which a local coordinate system was constructed. P1 is the reflective marker placed during calibration at the location where the beam was emitted from the laser. P0 is the point of intersection of the laser with the screen, calibrated with a reflective marker. P1 and P0 were used to determine the unit vector û along the laser beam in relation to the local coordinate system of the hat rigid body. Wall markers W1, W2, and W3 were used to calibrate the wall’s position in the global coordinate frame.
Figure 2.
Mean across subjects (±SEM) of the maximum joint excursions for all measured joint motions for quiet standing while visually fixating a point between two stationary targets and across cycles of tracking the moving target with the head; L5-S1 = joint between 5th lumbar and 1st sacral vertebrae, C7-T1 = joint between 7th cervical and 1st thoracic vertebrae. Cycles were defined as pseudo-cycles for quiet standing with the same average cycle time as with target tracking.
Figure 3.
Constant and variable error of targeting.
Upper Panels: Average across trials projection of the laser pointer (black dots) onto the screen at every 10 frames of tracking for one representative subject, after each cycle was normalized to 100 frames. Units are in meters (m). Lower Panels: Average across-subjects constant error (±SEM) and variable error (±SEM) of targeting with respect to the moving cursor are shown.
Figure 4.
Example of components of joint configuration variance across the tracking cycle.
Components of joint configuration variance per dimension of joint subspace (deg2) across the tracking cycle, related to stability of the 4-DOF CM position of a representative subject. Note the consistency of VUCM and VORT across the cycle, leading to using the average across the cycle for further statistical analyses.
Figure 5.
Average components of joint configuration variance.
Mean (±SEM) components of joint configuration variance per dimension of joint subspace, related to stability of the center of mass position based on a 6-joint (CMPOS-6DOF) and a 4-joint (CMPOS-4DOF) geometric model and a 2 segment model, and for head orientation with the target (HEADORI), all computed during quiet standing while visually fixating a point between two stationary targets and across cycles of tracking the moving target with the head. Significant differences between VUCM and VORT: *p<0.05; #p<0.005; ##p<0.001.
Figure 6.
Relative difference between components of joint configuration variance, overall and for individual joints.
Mean (±SEM) normalized variance difference between VUCM and VORT for quiet standing and target tracking tasks before (patterned bars) and after (solid bars) removal of covariance among the joints (COVAR REM) by randomization, displayed for each performance variable. Values close to 1.0 indicate that most of the joint variance was VUCM. The amount of decrease after removing joint covariance reflects the extent to which VUCM was due to multijoint coordination. Possible values range between −1.0 and 1.0. Only the positive range is shown for the upper two panels and lower left panel because no values fell below zero, unlike individual joint contributions. Lower right panel shows individual joint contributions based on both the 6-DOF CM model and HEADORI variables for the target tracking condition. Results for quiet standing differed minimally from those for target tracking.