synchronization problem?

[kurt manal]

Hi All,

the deadline is approaching quickly and I'm afraid I've run into a snag. If it's my fault then I apologize in advance. If not then this post will be of interest to others.

I've been working with the jw_tsgait_11 trial (the 2nd trial we are attempting to predict) and wonder if perhaps the data were not synchronized correctly. Plotting the vertical ground reaction forces for plates 3 and 4 with the vertical coordinates of the right and left heel markers indicates the left foot should be in contact with plate 3 and the right with plate 4. I believe the correct sequencing should be the right foot on plate 3 and the left on plate 4. Could someone kindly verify this and respond as soon as possbile given the looming deadline.

Thanks in advance,

Kurt

[Morten Lund]

Hi Kurt

I have observed this as well. However, it didn't occur to me that it could be caused by a synchronization problem.

My guess was that the markers were mirrored. So I simply swapped the left and right markers.

The link below shows a plot of "ground reaction forces" and "x heel marker" in the original tsgait_11 file. This is clearly wrong.

http://dl.dropbox.com/u/1683635/jw_tsga ... iginal.png

If I swapped left and right markers and made the plot again it seems to be ok.

http://dl.dropbox.com/u/1683635/jw_tsgait_11.png

Maybe BJ can confirm this...

/Thanks
Morten

[Thor Besier]

Hi All
apologies for not monitoring this thread sooner. BJ notified me this morning of this issue and I will look into it right away.

It surprises me that the trajectories and GRF are out of synch, since they were collected synchronously by the motion capture system. I found the common time point for heel strike (from the vertical GRF of FP 3) and then shifted the time base of the Eknee and EMG data to match. The marker trajectories and GRF data were then resampled with a spline, along with the Eknee data to match the capture frequency of the raw EMG data (which is why the GRF data are at 1000 Hz).

My initial response would be that the markers were labelled incorrectly for this trial...

more soon..

thor



and to be honest I did not check these, simply because they were collected by the same computer.

[Thor Besier]

Hi Again,
I dug up the digital video for this trial and concur with Kurt that the right foot is in contact with FP 3 and the left foot is in contact with FP 4.

It seems to me that the markers were labelled incorrectly for this trial. I am hoping this is the only one, but will have to do some checks to make sure.

Apologies for this!

Thor

[B.J. Fregly]

I checked with the student in my lab who has already processed all the trunk sway gait trials, and he confirmed for me that he had to rename all the left side markers as right side markers, and vice versa, to get ts_gait11 to process properly. He did not remember having to do this for any of the other ts_gait trials. So ts_gait10 should be fine.

So bottom line - markers from the right and left sides are definitely swapped for ts_gait11. My apologies that I did not realize this problem sooner.

If you simply swap the marker names between the left and right sides for the raw data as well as the synchronized data for ts_gait11, everything should work fine.

B.J. Fregly