I’m using AnyBody v8.2 to investigate lifting tasks based on MoCap data, focusing on how different marker protocols affect estimated spinal muscle activity. I’m using the ammr 3.2.0 Plug-in-gait_MultiTrial_StandingRef Model as a starting point and the GRF Prediction is activated.
Is it recommended for my application, to include passive structures for the spine for a more realistic approach? If so, would you recommend including spinal ligaments, disc stiffness or both? And for the disc stiffness, would a linear or nonlinear setting be more appropriate? I’ve observed, that without passive structures, the muscle activity profiles are relatively smooth over time. However, when I enable ligaments and nonlinear disc stiffness, the profiles show more spikes and fluctuations. Is this behavior expected? Are there strategies for dealing with it? Perhaps by increasing the cutoff frequency for filtering the marker data? So, what would be a reasonable cutoff frequency for a marker data sampling rate of 100 Hz?
The ligaments and discs will not affect optimized kinematics by default. But it is possible to include them into the optimization strategy: you would need to create additional “soft” drivers using potential energy formulation for elastic tissues. But my worry would be that the stiffness values are not truly patient-specific and you would not be getting “real” kinematics anyway. You would just introduce additional uncertainty around stiffness value and driver weight parameters.
Stepping a little back, if your focus is on varying marker protocol - I would closely inspect the difference in markers on thorax and upper body, since pelvis markers are quite reliable and typically have larger weights. This difference and consequent change in individual marker contributions to the objective function for kinematics optimization will explain the differences. I believe you use the rigid thorax - all the difference root in the position of the thorax.
I would not expect any sudden jumps in muscle activities due to adding discs and ligaments IF the predicted kinematics was previously smooth. Ligament and disc forces should be somewhat smooth unless we are observing some sudden accelerations.
Does this answer your question a little? Can you provide some plots and potentially more information about what you are doing? And how you approach “different marker protocols”?