Initial Values of Sholl Analysis


I reconstructed a neuron in FIJI and ran the Sholl Analysis by selecting the ctrl+shift+A function in the middle of the soma. However, at a radius of 0, the output file displays there are 2 crossings.

Am I reconstructing the soma incorrectly? I create 2 points on the edges of the soma to create a separate path for the soma. I use the ctrl+shift+A functon in the middle of this path,using the crtl function to navigate this path. I viewed the FIJI tutorial, where the origin of the path is the middle of the soma, and the sample image also displayed 2 intersections at a radius of 0.

Thank you for your help!

@tferr Just mentioning you here in case you didn’t see this!

1 Like

There are multiple ways of reconstructing the soma. Please have a look at what most anatomists have agreed on. I guess when Mark Longair wrote Simple Neurite Tracer he only considered the simplest definition of soma: a single point.

Since this is a rather poor strategy to define the soma, I consider it a bug in Simple Neurite Tracer. I promisse to have a look at it, and will try to correct it (hopefully I will find the time to do it fast). I filed an issue so it is not forgotten.

Meanwhile, you would have to either: 1) discard radii smaller than the soma radius, or 2) define the soma as a single point (i.e., a zero-length path – which you can create by generating a path in which starting point and end point are the same voxel). But I feel quite uncomfortable in recommending this latter option, as again, it is rather simplistic, and it will hurt your ability to use the tracings for electrotonic modeling.

1 Like

Thank you very much for your fast reply and explanation!

Just to mention that last week I proposed a fix for this.
No mater how you trace your somas, you will be able to exclude somatic traces from the analysis, using a new filtering option. It also makes possible to obtain profiles from tagged structures (such as branch points), which was not possible before.

The new fix will be available through the updater once it is revised and merged.

@ctrueden Just mentioning you here in case you didn’t see this! :slight_smile:

1 Like

I saw it, and I appreciate your efforts very much. I am aggressively avoiding reviewing PRs until I complete the ImageJ2 paper draft. I will try to get through my backlog as quickly as I can after that.

One option to speed things up would be for you to take on the responsibility of maintaining the Simple_Neurite_Tracer project, since it is currently unmaintained (i.e.: minimally maintained by me only). Of course, I completely understand if you do not want to commit to that, but it does fit in with your support and maintenance of Sholl Analysis. And you can commit to whichever team roles you have time for. If you became the SNT maintainer, you could merge your own PR here without guilt. :wink:

I guess it makes sense. Sure. At least for now. But I will let you know asa I can no longer commit to it. Good luck with the manuscript! We feel your pain!

Great! I added you as lead + maintainer (fiji/Simple_Neurite_Tracer@07d9154e). If you want to commit to any of the other roles (particularly debugger, reviewer or support), please add those to the POM as well. Pretty soon, the sidebar for each plugin page on the wiki will be autogenerated from these POMs, so it actually matters.

I also created a new “Simple Neurite Tracer” team in the fiji org, and invited you to join. Once you do, you should be able to merge your own PR if that’s what you want. Otherwise, if you want code review, just leave it for now and I will review as soon as I can.

Of course. These things do not have to be forever.

It is so close to finished. Just making figures now.