Problem with AnyGait

Main Forum
1

I try to use AnyGait in order to analyse some gait data. In total we have 9 subjects. For one subject AnyGait works perfectly, while for others I get an error. In most cases AnyGait stops in the first step, because of the KinSolver and the Initial conditions.

Project 11_G17 works (frames 918-1212). Project 11_G110 (frames 575-874) works too, however, in the last section of the movement, the upper body gets tilted back producing high loads in some muscles. Project 13 does not work. and I have the same or a similar problem an most of my other subjects/trials. I’m not quite sure how to improve my model to make it work and I’d appreciate your help.

Btw I have changed some of the marker names to get the model run with our marker set.

My AnyBody version is 5.3

Best regards,
Nathalie

2

Hi Nathalie,

first of all, good job to configure AnyGait to your needs, this is not an easy model. About your errors, the stick figure does, what the markers does, so if you look at Project 11_G110, you can see that the posterior markers are dropping out. That means they go to the origin in 0,0,0 and that leads to the pelvis moving down. Project 13 has the LTHI marker on the ground when loading the model, that shouldn’t be. If you out-comment the LTHI marker it runs through. I will look deeper why LTHI is not found.

Another point, please be aware that this is a public forum, so anybody can access it, so please make sure that you do not upload any confidential patient data, name etc…

3

Hi Amir,

thanks for your help.

2 more questions:
where can I out-comment the LTHI marker? do you have an idea why it is not found?

Concerning Project 11_G110: when I look at the c3d file in Vicon Nexus, the posterior marker isn’t dropping out until frame 879. (Also checked it with Mokka, it’s the same frame there). I entered in AnyGait frames 575-874, therefore I thought that there shouldn’t be a problem with the marker dropping out. Is there a shift/ difference between the frame number I see in Vicon Nexus and the one used by AnyGait?

Best regards,
Nathalie

4

Hi Nathalie,

  1. in the Drivers.any. If you switch from Project View to the Model tree view and double click on an object in the Model View, the Model Tree will jump to the object and you can right click on it “open in script” this will bring you to the location of e.g. the markers…

  2. There is a filter of the markers included that will affect the curve of the markers, please compare in the AnyChart:
    Filtered Marker position
    Main.ModelSetup.DynamicDataSet.Points.Markers.RPSI.PosFiltered
    original Marker position
    Main.ModelSetup.DynamicDataSet.Points.Markers.RPSI.Pos

that means the marker drop out affects the position of the marker 5-10 frames before…
PSI markers are very important, if it would be another one you could get away with out-commenting it.

5

Hi Amir,

thanks a lot for your previous help, I was able to run AnyGait with most of my data. I still have problem on 3 projects though… I can’t quite figure out why they aren’t working.
In all 3 projects (attached) Kinematic analysis fails in time step 0.

I’d appreciate if one of you could have a look at the data and might give me some feedback on what I have to change to get in running.

One more question:
For one subject, there is a mistake in the computation of the muscle force curves (figure attached). I thought that maybe markers were dropping out causing the the ‘jumps’ in the curve, however, I couldn’t see markers dropping out in my c3d file. Do you have any suggestions what else could be the reason? My the force plate data?

Thanks for your help,
Nathalie

6

Hi Nathalie,

I was going through the 3 simulations, but so far I couldn’t find the error. the 15_re runs, when I use a different static file. But I don’t know yet why???

Why do you think that there is a mistake in the muscle force simulation. What is this graph showing? Is it the sum?

7

Hi Amir,

thanks for your reply. I’ll try to use a different static trial then. Which one did you use to run it?

Yes it is the sum. I thought there might be a mistake, as I get a ‘smooth’ curve for the rest of my subjects (see attached figure). In the figure presented before, it seemed to me like the development of the peak muscle force of the the injured muscle is interrupted.
However, I have hardly any experience with AnyBody data so far, so I’d appreciate your feedback, if this very well could be a correct computation of the gastrocnemii muscle force.

Thanks,
Nathalie

8

Hi,

I don’t think that there is a computational error, because everything is exactly the same as for the other subjects.
It looks to me that you have maybe one branch of the gastroc acting earlier in the gait cycle, resulting in the first peak of the curve. I would suggest that you look into the individual muscle forces for this subject. This effect is caused by the kinematics (the way this subject walks) and there should be a reason why AnyBody predicts that the gastroc is active at this stage.

I used a static trial form Khalif_11 to get 15_re running, however, I couldn’t get 13 running (rechts & links). This seems little strange, I’m not sure why this is the case yet.

9

thanks for the explanation! I will have a look into it!