I am currently working to calculate the dynamics of the human joints during motion. For the inertial properties I am using body segment inertial parameters (specifically the ones reported in: "Adjustments to McConville et al. and Young et al. body segment inertial parameters", R. Dumas et. al).

Now my problem is that the coordinate frames of the joint centers on my model do not have the same orientation as the ones presented in the paper .

Is it correct to calculate the tensors of inertia and then rotate the matrix appropriately to match the coordinate frames or it is better to change my kinematic model completely and not meddle with the tensors?
Fairly sure both methods are valid but has anyone attempted that before?

Thanks in advance.