Energy cost prediction

[Sauvik Das Gupta]

Hello,

I have used OpenCap to collect some pilot data and have been able to predict standard gait parameters, kinematics and kinetics for various movements (like squats, counter movement jumps, sit-to-stand and treadmill walking, etc.). I have mainly tried to solve the torque and muscle-driven dynamic simulations, from the example scripts provided, to solve for the above problems.

I am now interested to try to predict the metabolic/energetic cost of these activities. That way, I can try to maybe solve the problem of predicting metabolic cost of these movements from videos alone (collected through OpenCap).

Can someone please point me to the right direction? Please also let me know if this is not the correct forum for this question and if I need to post it somewhere else.

Thanks a lot!

[Antoine Falisse]

Hi Sauvik,

There is no support for metabolic/energetic cost estimation at the moment. However, the code you used to generate the simulations is very similar to the one here https://github.com/antoinefalisse/predsim_tutorial used to generate predictive simulations of movement. In that code, I computed the metabolic cost of transport, you could use that as inspiration: https://github.com/antoinefalisse/preds ... n.py#L1627

Best,
Antoine

[Sauvik Das Gupta]

Hi Antoine,

Thank you very much for the guidance to the PredSim tutorial. I will try to take a detailed look at it.

I quickly took a glance at the main.py code and saw that you may have probably used the Bhargava model to do the Met. cost predictions. Can we only use the Bhargava model or maybe replace it with some other muscle-space models too, like the Umberger (probably used in OpenSim metabolic calculators)/Lichtwark/Minetti/Margaria/Houdijk models?

Additionally, can the KIMR15 model be also used as a joint-space model?

Just wanted to know your opinion!

Thanks,
Sauvik

[Antoine Falisse]

Hi,

You could use whatever model you like. I think I only have python implementation of the bhargava one. I have matlab implementations of some other ones though: https://github.com/antoinefalisse/3dpre ... olicEnergy.

> Additionally, can the KIMR15 model be also used as a joint-space model?
I am not familiar with that.

Best,
Antoine

[Sauvik Das Gupta]

Hi Antoine,

Ok, I understand. Thanks very much for the GitHub link to the other muscle models. I will take a look.

Thanks,
Sauvik