Scale tool leaving markers behind
Posted: Mon Aug 07, 2023 2:31 pm
Hello,
I imported marker data from a .c3d file using the c3dExport.m and osimC3D.m files. I had to change the osimC3.m file slightly because my marker set had a bunch of calculated joint center and bony landmark trajectories in it that I needed to remove. I did so using osimTableToStruct to get the data in a format I could remove columns from and then put everything back together using TableFromStruct and addTableMetaDataString. Now, I'm working on scaling my markers to the Gait2392 model. I have a set of 20 markers. I adjusted the coordinates of the model so that it looks like the pose of the person I'm scaling to and adjusted the positions of the markers so they look like where the markers are on the person I'm scaling to. Then, I set up my scale tool. I loaded in the marker set, the marker data for measurements, and the marker data for static pose. I specified my participant's mass, and selected preserve mass distribution during scale. Then I set my scale factors:
pelvis, uniform, LASI/RASI
femur l/r, uniform, lateral knee and trochanter marker pair
tibia l/r, uniform, lateral ankle and lateral knee marker pair
talus l/r, calcn l/r and toes l/r, uniform, heel and toe marker pair
torso unassigned
I set my weights accordingly:
Bony landmarks 5 (LASI/RASI, RKNE/LKNE, RANK/LANK, RPSI/LPSI)
Soft spots 1 (RTHI/LTHI, RTIB/LTIB)
Non-joint bony landmarks 1 (RHEE/LHEE, RTOE/LTOE, LGreatTroch/RGreatTroch, LIliacCrest/RIliacCrest)
Then, I ran the scale tool. My LGreatTroch/RGreatTroch, LIliacCrest/RIliacCrest, and RTIB/LTIB markers do not snap to the body of the model at all. In fact, they hover in space far away from it. I have tried increasing their weights to 6, un-checking "preserve mass distribution during scale" since the participant was wearing an exoskeleton for the data collection, and changing my scale factors such that the hip z-direction is defined by the RPSI/RASI markers and the hip y-direction is defined by the LGreatTroch/LIliacCrest markers. None of these strategies have worked for me.
I am looking for advice on why those 6 markers are not attached to the model after scaling. Any help would be greatly appreciated.
Thanks,
Emily
I imported marker data from a .c3d file using the c3dExport.m and osimC3D.m files. I had to change the osimC3.m file slightly because my marker set had a bunch of calculated joint center and bony landmark trajectories in it that I needed to remove. I did so using osimTableToStruct to get the data in a format I could remove columns from and then put everything back together using TableFromStruct and addTableMetaDataString. Now, I'm working on scaling my markers to the Gait2392 model. I have a set of 20 markers. I adjusted the coordinates of the model so that it looks like the pose of the person I'm scaling to and adjusted the positions of the markers so they look like where the markers are on the person I'm scaling to. Then, I set up my scale tool. I loaded in the marker set, the marker data for measurements, and the marker data for static pose. I specified my participant's mass, and selected preserve mass distribution during scale. Then I set my scale factors:
pelvis, uniform, LASI/RASI
femur l/r, uniform, lateral knee and trochanter marker pair
tibia l/r, uniform, lateral ankle and lateral knee marker pair
talus l/r, calcn l/r and toes l/r, uniform, heel and toe marker pair
torso unassigned
I set my weights accordingly:
Bony landmarks 5 (LASI/RASI, RKNE/LKNE, RANK/LANK, RPSI/LPSI)
Soft spots 1 (RTHI/LTHI, RTIB/LTIB)
Non-joint bony landmarks 1 (RHEE/LHEE, RTOE/LTOE, LGreatTroch/RGreatTroch, LIliacCrest/RIliacCrest)
Then, I ran the scale tool. My LGreatTroch/RGreatTroch, LIliacCrest/RIliacCrest, and RTIB/LTIB markers do not snap to the body of the model at all. In fact, they hover in space far away from it. I have tried increasing their weights to 6, un-checking "preserve mass distribution during scale" since the participant was wearing an exoskeleton for the data collection, and changing my scale factors such that the hip z-direction is defined by the RPSI/RASI markers and the hip y-direction is defined by the LGreatTroch/LIliacCrest markers. None of these strategies have worked for me.
I am looking for advice on why those 6 markers are not attached to the model after scaling. Any help would be greatly appreciated.
Thanks,
Emily