Weighted streamline-fixel attribution

[Lestropie]

Relevant manuscript:
https://www.frontiersin.org/journals/neuroscience/articles/10.3389/fnins.2023.1199568/full

I'm not sure about the fully continuous expression that they propose there.

With #2657, there will be additional information utilised for streamline-fixel mapping when streamlines have come from FOD segmentation, being the dixel mask per fixel. So as long as the tangent of the streamline-voxel intersection is a part of the lobe that formed that fixel, it will be assigned. But there still needs to be a decision made where that intersection doesn't intersect with any preserved fixel. In that scenario, a weighted attribution of that streamline segment to multiple fixels makes more sense.