User report of model slightly underestimating CSA

[joshuacwnewton]

From the SCT forum:

As a neuro-radiologist, my opinion could be helpful for developing a better algorithm. Although current segmentation is good enough to use, but to me , at visual inspection, segmentation result by seg_lumbar_sc_t2w algorithm tends to underestimate CSA a little bit.

I'm not sure what actionable steps can be taken here (e.g. reaching out to this user, potential refinements/collaboration, etc.) but I wanted to make the connection just in case. :)