SNT-Sholl Analysis Bug (Following update?)

I have been using the SNT plugin extensively to trace/analyze cell morphologies with no issues for months now. Yesterday, I followed a prompt for a simple software update and then proceeded to use the plugin with no issue (tracing functions and other reconstruction viewing functions work). However, when I attempt to perform sholl analysis with the same settings that I had used on all previous cells, the console shows that “Sholl Analysis starting…” with the progress bar registering that it is “obtaining profile…”, but then nothing occurs. It is as if the Sholl analysis is not working at all (the program sits at this stage for hours with no progress). Is there any idea about how to troubleshoot? Thank you-Luke
–For reference, I’m using ImageJ 2.0.0-rc-71; Java 1.8.0_202; SNT 3.1.103; Sholl 4.0.1

Welcome to the forum, @lukenunnelly! This is when using the Analysis > Sholl Analysis… menu option from the SNT dialog to measure tracings? Unfortunately, I am not able to reproduce this behavior on Windows 10 using the same plugin versions. Which platform are you on? Also, it would be helpful to know if cells that you were able to analyze successfully in the past exhibit the same behavior with the current version. If that is not the case, there could be an issue with the reconstruction itself, which I could take a look at if you don’t mind sending it.

Hey! Reconstructions that worked with previous versions no longer work following the update. I’m using this on IOS Mojave. What’s the best way to share reconstructions so you can take a look at it?

I see, in that case it is likely not reconstruction related…pinging @tferr

Hi @lukenunnelly, I also cannot reproduce it on ubuntu.
version 4.0.1 implements only an option to save all files into a common directory (this was mainly to facilitate the saving of tables). I don’t see any other major changes that could cause this – we could revert to v.4.0.0 if needed --. I’m wondering if something else got updated in your installation that is causing the issue. Could you please post the contents of Plugins> Debug> System Information (bottom of the plugins menu)?

Also, just to be sure, the current menu entries in Analysis> Sholl> are:

  1. Sholl Analysis (From Image)…
  2. Sholl Analysis (From Tracings)…
  3. Deprecated > Sholl Analysis…
  4. Deprecated > Sholl Analysis (Tracings)…

(the deprecated entries refer to the outdated IJ1 plugins, currently functional but no longer maintained). Which command are you evoking ?

On sharing reconstructions, you can post them here by appending .txt to the file extension.

Hello,
I had the same problem and it seemed to be resolved in v.4.0.1. Once filaments are traced, select them all and go to Edit -> Merge all Primary Paths into Shared Root…that seemed to fix the infinite hangup at Analyses Started, at least for me.

Good luck. thank you to @tferr for fixing the bug!

@folive, thanks for the clue. I can finally reproduce this issue. It does happen when the specified file contains multiple rooted structures (i.e., multiple disconnected primary paths). The software expects a file of a single rooted structure with no loops. When it finds multiple disconnected paths, it does not know what to do with it and hangs. I had never noticed it before, because all this time I had never generated such a file (hence the importance of sharing the file-triggering an error: it helps us reproducing the bug).

@folive, @lukenunnelly, I opened an issue so we don’t forget about it. How would you envision the handling of such files?

  1. Abort parsing and display an error message
  2. Include a drop-down menu in the prompt that would allow you to choose which structures/arbors to parse. This would be perhaps the best, but may slow down significantly the loading of the file.

I’m no expert but perhaps this occurs when the primary paths do not start exactly at the center point of reference located in the cell body (i.e., start point of first primary path). Even if subsequent paths start a little bit off from the original point, even just a few pixels, perhaps they are marked as disconnected. Could be due to improper zoom level or not starting new path at nearest node. Just a hunch. Thanks for all your help!

Foster

Thank you both for our help! Both of these fixes have helped to progress my analysis again! I think just an error message with a suggestion box/troubleshoot would help avoid getting completely stuck. Thank you again!

Luke

I agree with @lukenunnelly. An error message would be good so the user is alerted, rather than having it hang indefinitely. thank you for all your help!

I did spend some time trying to find a way to automatically handled files with multiple cells/disconnected paths. It is very hard to predict every edge case, so I’m putting that on hold for now. I did add an error message. Will be available with the next release (in a week or so).

Thanks for your help!