Issues with Sholl Analysis in Simple Neurite Tracer

Hello, I am a student and new to the forum community. I am having some difficulties when it comes to performing Sholl analyses in the Simple Neurite Tracer program. I have traced select astrocytes using an .AVI file of a z-stack I took on a confocal microscope. I have selected the center of analysis by highlighting the first path and using CTRL+SHIFT to follow the path to the desired center of analysis. I then go ahead and use the CTRL+SHIFT+ALT prompt to initialize the Sholl analysis. All the paths traced are selected in the analysis, and I step a radius step size to a desired 3.65 pixels. The analysis runs perfectly fine, however, I am getting variable results (i.e. number of intersections). I have included screenshots to roughly demonstrate my process and the parameters I am setting. Is it an issue of my images having a lot of background that vary from slice to slice in the Z-stack? Is it possible that some of the processes extends tangentially to the sampling circles, making it so that a single process is contributing to multiple intersections? Would anyone have suggestions to help me achieve consistent and accurate measurements?

Capture4 Capture5

Thank you kindly,
Jan

@jkocur

I believe @tferr is the best person to assist you in this issue. But perhaps also share an original image file here or via a link to a file sharing site. Might be easier for folks to test things.

@etadobson thank you for the advice! I think I found a solution to my problem by conducting the following: Analyze > Sholl > Sholl Analysis (Tracings) and performing the aforementioned analysis on just the loaded tracings themselves. This has yielded accurate and consistent results thus far. I think the issue was in SNT the sholl analysis was inferring the number of primary branches and the subsequent intersections based on the images, which contained bits of background noise, ultimately skewing results.

1 Like