I'm wondering how, if even possible, rotationRate data can be interpreted for the turn portion (first few seconds) of resting data files. When each of the three axes are graphed, there appears to be no distinct direction of turn because positive and negative values for rotationRate negate each other. Mathematically, this should indicate no net turn. I have yet to reach an explanation for this phenomenon. Any insight or clarification on this topic would be greatly appreciated! Thanks, Matthew

Created by Matthew Tsai matthewtsai
Hi Mathew: The rotation rates are rates not actual degrees. If you want to see the rotation you have to integrate the rates. For example if you look at the first recordId (704fda87-91c7-4c67-b520-b8e189f7f7f7) from the training table (syn10146553) and plot the rotation rates from the rest you should see something that oscillates back and forth a lot in the beginning (Figure 1). But if you integrate across the first 300 samples you should see that the person is rotating through 180 degrees in the y direction with some wobbeling back and forth in the x an z direction (second figure) Figure 1|Figure 2 ${imageLink?synapseId=syn10791535&align=Left&scale=25&responsive=false}| ${imageLink?synapseId=syn10791534&align=Right&scale=25&responsive=false}

Interpreting rotationRate data during subject turn page is loading…