Problem with calibration data

Provide a comprehensive data set with associated models that will enable researchers to validate musculoskeletal model estimates of muscle and joint contact forces in the knee.
POST REPLY
User avatar
Morten Lund
Posts: 6
Joined: Thu Oct 01, 2009 12:11 am

Problem with calibration data

Post by Morten Lund » Sat Dec 26, 2009 2:09 am

Hi

I stumbled on something in the synchronized data which I can't make any sense of.

In "jw_2legsquat_1" the medial and lateral contact force seems to highest when the person is standing upright and decrease during squatting. This seems to counter intuitive.

Here is a plot of the approx knee angle and the contact forces, so you can see what I mean:
http://dl.dropbox.com/u/1683635/jw_2legSquat_1.pdf

Is there a problem with synchronization between eknee forces and and mocap data?

Regards
Morten

BTW. If anyone have found other oddities in the data set, please share them.
Top
User avatar
Thor Besier
Posts: 7
Joined: Tue Aug 29, 2006 2:07 pm

RE: Problem with calibration data

Post by Thor Besier » Thu Jan 14, 2010 12:02 pm

Hi Morten
Sorry for not following up on this thread earlier. The squat trial was a tricky one to synchronise as our common synch signal was actually FP 3 and I only realised afterward that our subject was standing on FP's 1 and 2. The only common signal for me to synchronise was the EMG signal from vastus lateralis, which was our other common signal.

To do this, I low pass filtered the EMG signal from both sources and then performed a cross-correlation to get the synch point. What I am afraid of is that I obtained the minimum value of the cross correlation and not the maximum for this trial, in which case the two signals would be 'out of phase', as you have shown.

I will dig up my code now and take a look.

Again, I apologise for not picking this up earlier.

Thor
Top
POST REPLY