Different muscle activation depending on analysis start

Dear AnyBody Team,

while looking at some MoCap running data, I encountered some difficulties.

Usually I run the AnyBody analysis 5 frames before my event of intereset until 5 frames after the second event of interest (during running these events would be heel strike to toe off, or second heel strike).

Since I did not get muscle activities as reported in the literature (mentioned in another post), I ran the whole file, which started with a heelstrike and toe off not on the forceplate before coming to heelstrike and toe off on the force plate.

I’d expect small differences in calculation of muscle activity depending on when I start my analysis. However, I found huge differences. I have attached 3 figures of 3 subjects showing two different conditions (red and blue) of vastus lateralis muscle activation. On the x-axis it is frames, so the blue curve showing the peak in the first 60 frames is from the file where I only ran the analysis from HS to TOE +/- 5frames, and the curve showing the peak at the end of the dataset is from the complete file.

I couldn’t check walking data we have caputered, since we usually only have files starting a couple frames before HS. Might those huge changes in muscle activation calculations be due to the movement (running)?

I have also checked one subject for changes in muscle activation when starting the analysis 15frames before HS. Changes occured, but were minor, especially compared to those presented in the figures.

I haven’t changed anything besides the length of the file for analysis. Can you explain me where those differences come from? Since different conditions behave differently, too?

I’d appreciate your help,

it is tough to say anything without seeing the model itself. There are many reasons for strange behavior…

Hi Amir,

I have uploaded my model in the Debug section: http://forum.anyscript.org/showthread.php?t=3779

Thanks for your help,