steps of including EMG data into AnyBody?

Hi, AnyBody,

I plan to include EMG experiment data into my AnyBody model.
(1) Is the EMG data enough to drive the model to move?
(2) Is the .c3d file required by AnyBody or without?
(3) Do I have to convert .c3d to .any file, using .c3d2any.exe tool?
(4) What is GaitApplication2.exe working for?

Can you please tell me the detailed steps that I can follow to include EMG (or c3d) into AnyBody?

(5) If I have angular kinematics data, can I import the data into AnyBody to drive the model to move?
What other steps that I need to follow to include angular kinematics data as AnyBody input?

Sorry for those many questions. I was told here is the primary place asking for help.

Thank you!

Cathy

Hi Cathy,
to dirve the model with EMG data is not really possible. The way to go is to use the C3d motion and force data. Please have a look at the tutorial:http://www.anybodytech.com/fileadmin/AnyBody/Docs/Tutorials/Making_things_move/lesson2.html I guess this will explain all you questions. There are also two recordings of recent webcasts available. John Rasmussen explained how to use the data. http://www.anybodytech.com/index.php?id=199
Please have a look, and come back if you have any further questions.
Best regards,
Sebastian

Sebastian,

Thanks for your reply! I will read the tutorial once again. But I have one more question here:

I have a BIOSYN FAB system, which will give me realtime angle, force, torque, and power data set in .csv file format. The FAB system is with 13 sensors on a full body: Head, Left/Right upper arm,Forearm, thigh, calf, and foot pressure, Trunk and Pelvis.

After I convert the .csv into .c3d format, will it be the (same) way to move the AnyBody model? If the angular kinematics data from FAB enough for the model to move?

I read other threads about using joint angles as input, using AnyKinEqInterpolDriver function. Will that be the same as what I want to do with FAB system? Does AnyBody have such example?

Because you know best about AnyBody, I want to know your thoughts about whether my idea is feasible or not.

Thank you!

Cathy

Hi Cathy,
you can use the joint angles directly with AnyKinEqInterpolDriver functions, thats right. I don’t know what is in your C3d file. What we basically use is the positions of the markers. Maybe in your case its easier to use the joint angles directly, e.g. with individual textfiles for each angle variation.
I don’t know the Biosyn system, but I would be very careful with the data. You know we are are doing inverse dynamics and everything depends on proper kinematic data. We have very limited experience with such devices, usually we use data from full blown gait labs, but in the few cases we have touched on this we found that the data was simply not accurate enough. The least you have to do is to apply filters to get smooth data, there is an example in the tips and tricks section of the Anyscript page.
Best regards,
Sebastian

Sebastian,

Thanks for your kind reply. I will try to input joint angles directly according to your suggestion.

With my best,
Cathy

Hi, Sebastian,

Sorry to bother you again. After my discussion with my advisor, we have such a question as bellow about EMG. Will you please answer me this question in detail?

We have C3d motion and GRF force data and also EMG data from some of the involved muscles.

Based on my reading I understood that I can run my AnyBody model using motion and GRF data and can use the EMG data indirectly by doing a correlation between the force and EMG data.

Is there any other way I can use my EMG data more directly by incorporating in the model like motion and force data?

Thank you so much!

Cathy

Hi Cathy,
no, there is no direct way. If you know the relationship between force and EMG signal then you could apply forces to the model along the line of action of the muscles. But this is a very uncommon approach and not at all recommended. The usual workflow is: Create the AnyBody model and run it with marker and force plate data. Analyze the model and use EMG data to validate the results. Have a look at this paper:
deZee, M.; Dalstra, M.; Cattaneo, P. M.; Rasmussen, J.; Svensson, P. & Melsen, B. (2007), ‘Validation of a musculo-skeletal model of the mandible and its application to mandibular distraction osteogenesis.’, J Biomech 40(6), 1192–1201.
This shows how the EMG data can be used for validation purposes.
Best regards,
Sebastian