This has been done at the segmented bone’s surface. May be you could grow the segment by 1 mm at least, apply Joint smoothing before converting to a model. This would account for the thickness of the cartilage, which is itself an invisible structure on a CT scan.
This is an interesting approach; thanks for introducing the Dynamic Modeler module.
By any chance, is there any option to interpolate the segmentation? At the moment, I have 31 slices and interpolating the image doesn’t fix these steps, I have to do segmentation in all interpolated images.