Inconsistency Graichen et al. 2000, check is isb_like from its initial definition about spartacus-shoulder-kinematic-dataset

[fmoissenet]

I was not able to find again the humerus axes definitions in the paper, but I guess that +anteroposterior (Humerus_X) should be replaced by -anteroposterior. Debate displaced in #42.

[ANaaim]

I do not agree @fmoissenet. Here the data might be globally oriented as the ISB but not be ISB (if the axis are not constructed with the anatomical landmark from ISB). We should not change the data in the table.

However, as in this article the thoracohumeral angle is controlled by a an operator and that the angle measured (scapulothoracic and sternoclavicular angle) do not imply the humerus we do not care about the humerus.

[Ipuch]

Then, we are in a tricky case...

1. Are the axis really different from ISB ? If very close, consider ISB = yes
2. Other possibility, remove data info from humerus is not used.
3. If not, we should either attribute a risk or something to say that despite our data, this is not ISB...

[Ipuch]

-> Test to add verify if everything ISB but not is_isb then correctable == false and no correction available.
This is a is_isb_like from the beginning. edit to #44

[Ipuch]

Anyway, we don't care since the humerothoracic angle is driven by an operator and not computed from rotation matrices.