In our first IRB approved study, investigators had young adult men and women perform balance assessments.  Investigators used three modalities to simultaneously gather data as subjects performed the Balance Error Scoring System (BESS) protocol.  In addition to BESS, investigators measured balance with a force plate and our MimicBalance! system.  Four assessments were conducted with 5 minutes rest between assessments. 

Correlation coefficients for the force plate-based measures compared to MimicBalance! measures on single leg, firm surface assessments were 0.755.   The CC's were high enough to be of interest, particularly given that Mimic captures motion that force plates do not.  As explained below, Mimic captures motion that the force plates miss!

The graphics compare the force plate data (the blue trace) with the Mimic data (the red trace).  Seven seconds into the assessment, the subject loses balance, steps to the floor, and then recovers.  The M-L graphic reveals motion clearly captured and reflected in the Mimic trace, but absent from the force plate trace.  In addition, the blue trace shows greater amplitude than the Mimic trace.

We hypothesize that at least three factors contribute to the difference.  First, ‘active ankles’ effect rapid weight shifts by the subject.  Second, the Mimic trace reflects averaging of multiple skeletal points data that has a smoothing effect.  Third, the force plate calculates center-of-pressure while Mimic calculates center-of-gravity.  As a result, when the subject loses balance and places one foot on the floor, the magnitude of the pressure decreases, but the locus remains relatively constant.  The subject's center-of-gravity however, drifts considerably. 

Therefore, the graphics illustrate reasonably high CC's, but also why Mimic measures are distinguished from force plate based measures.  Mimic clearly captures motion that force plates miss, including errors that according to the manufacturers invalidate the force plate measures all together!