Hi,
I’ve been doing a lot of reading on these forums and managed to piece together a lot, however I’m still having some difficulty solving my problem. My aim is to extract the centerline of a vessel tree in order to compute various metrics for each branch/section of the tree ( e,g, min/max/avg radius/curvature/torsion/tortuosity, volume/surface area of branch etc).
In most cases, the auto end points are working well (and made better from this How to obtain the complete center lines using VMTK? - #11 by perecanals ). In most cases however, the centerline computation is really poor, even with doing oversampling from here (How to obtain the complete center lines using VMTK? - #4 by lassoan). FWIW I know from postprocessing that my segmentation method is completely connected. see img for centerlines (in light green):
Regardless, in most cases the network calculation is good (see below for another example), so I am not too bothered about centerlines provided the network can be used to calculate the metrics I need. My question is is this possible? Would I be missing out on anything by not using centerlines?
I’m sorry if I’m missing something, maybe there is something within the vmtk package that already does this that i’m missing?
