CMC analysis of "Upper Extremity Dynamic Model"

[Hamidreza Abbaspourazad]

I am trying to get arm muscles activation from some recorded joint angles using OpenSim 4.0. I transfer my joint angles to the corresponding motion file as explained in thishttps://simtk-confluence.stanford.edu:8 ... ith+Matlab page. I use the arm model from here: https://simtk.org/projects/upexdyn. I am facing few issues and would appreciate it if you could help me with them:

1. First, when I load the arm model in GUI, I immediately get this warning:

"Coordinate 'deviation' (and 'flexion') was labeled as 'Rotational' but was found to be 'Coupled' based on joint definition. "

Having that in mind, when I import all of joints angles in Radian (compute it in radian in matlab and transform to .mot as I explained above), all other joint angles values in the coordinate toolbar will be automatically transformed and shown in degree (even though I input the Radian numbers) except for flexion and deviation which won't transform, and the movement visualization has a lot of variation in wrist which I assume is for this inconsistency. Therefore, I decided to compute all joints in Radian except for 'deviation' and 'flexion' in degree. This solved the movement problem, meaning that the wrist movement now visually makes sense. BUT, for some reason the 'deviation' and 'flexion' range in the coordinate toolbar is in Rad (unlike other joints which the range shown is in degrees). How do I fix this inconsistency?

Capture_GUI.PNG (17.45 KiB) Viewed 219 times

. This inconsistency made some problems in CMC analysis unless I lock wrist deviation and flexion as 0 for the whole course, which does not eventually desirable as I need the muscles involved for wrist movement too.

2. I have difficulties understanding the CMC tool. Can I use the same CMC setup.xml which is used for the CMC tutorial of the Upper Extremity model in my case too? If no, what things I have to keep/change in the .xml file to get good activation estimations?

3. If I have a 5min-long recording, with sampling ratio of 10 ms, I can use the GUI for muscle activation estimation (It looks like it takes extremely long)? if not, are there any ways like using Matlab to do it automatically and not through GUI?

Thank you so much for your help, I greatly appreciate your time for responding to this.