Scaling problem - RASI & LASIS issue

Provide easy-to-use, extensible software for modeling, simulating, controlling, and analyzing the neuromusculoskeletal system.
POST REPLY
User avatar
Karelia Tecante
Posts: 7
Joined: Thu Jun 14, 2012 10:35 am

Scaling problem - RASI & LASIS issue

Post by Karelia Tecante » Mon Aug 26, 2013 4:29 am

Hello

I am having issues with the scaling process and hopefully some of you might have more experience with this particular problem!

The resultant scaling gives me a very big hip (which might also be causing some foot scaling problems) and I manually make an adjustment of virtual markers location to better match the experimental markers. In some cases the size of the hip seems better but the thighs scaling seems incorrect. I have played quite a lot with the virtual markers location and I still think the scaling is not good.

My scaling problem occurs with high BMI (body mass index) subjects. I believe the problem comes from MOCAP (Vicon nexus system) since the RASIS and LASIS can not be seen by the cameras I am forced to move these markers more towards the side so that they can be visible by the cameras . In such cases I can measure the real inter ASIS distance of the patient and write it in vicon during MOCAP. Is there a way to do something similar in Opensim? can I input the interASIS somehow in OpenSim? Has someone had any smiliar problems and could somehow resolve it?

Note: I am using the PIG model (VM and EM)

I would absolutely appreciate any advice!
Thanks
Attachments
Scaled model.png
Scaled  model 2.png
Scaled model markers corrected.png
This is the resultant model once I have corrected the RASIS and ASIS. The femur seems wrong.

User avatar
David John Saxby
Posts: 83
Joined: Mon May 09, 2011 8:39 pm

Re: Scaling problem - RASI & LASIS issue

Post by David John Saxby » Wed Aug 28, 2013 9:33 pm

Hi,

You are describing a common problem. Scaling is performed by applying a scale factor to the body by dimension or uniformly across dimensions. You could determine the virtual marker positions in your generic/unscaled model, and compare this with your experimentally measured inter-ASIS to determine the scale factor offline.

s = virtual/exp

Then take this scale factor and apply to the body/dimension manually. In ScaleTool there is an option to apply manual scale factors.

David

User avatar
Karelia Tecante
Posts: 7
Joined: Thu Jun 14, 2012 10:35 am

Re: Scaling problem - RASI & LASIS issue

Post by Karelia Tecante » Fri Aug 30, 2013 8:03 am

Hello,

Thank you a lot for the input! I managed to improve my scaled model!

One more question...I have the same subject pre and post surgery and I am trying to use the same scaled model I used for pre in post (6 months after). I believe I only need to adjust the markers model (of the post motion capture) in the scale tool. For this I use the neutral trial, however Opensim seems to not like it and crashes. Are my thoughts correct? maybe my approach is wrong. Any ideas why this might be happening?

I attach an image of the error.

Thanks a lot
KT
Attachments
Adjust markers.jpg

User avatar
David John Saxby
Posts: 83
Joined: Mon May 09, 2011 8:39 pm

Re: Scaling problem - RASI & LASIS issue

Post by David John Saxby » Sat Aug 31, 2013 9:55 pm

Hi KT,

Regarding the java error I am not totally sure. I believe though that you must tick the scale model box for the model adjuster to run. I think this is due to some requirements in the cost function that the adjust markers uses. The Osim developers would know why.

Its not a problem though, you can enable the scale tool and simply set the scale factors to 1 that way the model dimensions are not affected but the marker positions are adjusted.

On a separate note, have you considered the validity of using neutral position + adjust model markers?

David

POST REPLY