How to save module display window


#1

Hi all,

I am having trouble with saving module display window. I use the SaveImage module, where there is a “Select the module display window to save” option in the supporting document. However, in the settings, the only five image types are Image, Mask, Cropping, Movie, and Object, and I cannot find module window anywhere. Could anyone tell me how to save the module display window since I want to save it in each cycle and I cannot do it manually.

Thanks in advance!


#2

Hi,

I’ll you refer you to this: viewtopic.php?f=14&t=4133&p=12426&hilit=save+module+window#p12434

Cheers,
David


#3

Hi,

Thanks a lot for the information!

Sincerely,
Siheng


#4

Hi David,

I need to save module windows (specifically, the subplot called Neuron graph from the MeasureNeurons module), but I can’t find the way to do it. That’s how I ended up in this post, but the link you provided to sihengchen doesn’t seem to be working anymore. The only way I found to save the Neuron graph is right-clicking it and then chooking “Save subplot”, but I need to do it in a high-throughput way.

Additionally, it would be great if I could overlap the Neuron graph lines with the original image (to be able to show what the pipeline is picking and what not, without having to compare the two images), but I don’t know if this is asking too much.

Many thanks for your help,
Adrià


#5

Hi,
You can save the branchpoint image by turning on Retain the branchpoint image? option in MeasureNeurons and then adding an additional SaveImages module to your pipeline. You can add images using the ImageMath module, again followed by a SaveImages module if you want to save it, not just view it.


#6

Hi Beth,

Many thanks for your reply. The image I would like to keep is not the Branchpoint image but the Neuron graph (the “subplot” on the right of the MeasureNeurons module view window, see example attached), precisely with the Linear option in the Interpolation parameter (as shown in the image). And this is the image that I would like to add on the original image with the ImageMath module. I’m afraid maybe it is not possible to get this image unless I right-click and then save every image one by one?

Cheers,
Adrià


#7

My apologies for the misunderstanding- yes, that seems like a problem. I’ve opened a GitHub issue to ask our software engineers to add an option to allow you to save that image.


#8

That would be awesome, thank you!

I’ve got another question related to this “How to save module display window” issue: I’m also interested in saving a variation of the original image on which all the processing is done in order to run the MeasureNeuron module; concretely, I’d like to save a display window image after applying a FilterObjects module and with the Image contrast set as “Log normalized” (see attached screenshots image -I’m interested in the image below). Is there any way of doing this automatically?

Regards,
Adrià


#9

There isn’t a way to do save random module windows or sub-windows, no- you can take a look at this old GitHub issue where the team discussed it, but apparently it would have been pretty challenging to do.

If you want to approximate this, you could do a combination of RescaleIntensity (there’s no log option there, but scaling from 0-0.1 or something like that usually does the trick) and then OverlayOutlines- because of the way OverlayOutlines works, if you draw two sets of outlines the second set of outlines will be drawn over the first set, making the first set invisible. Therefore if you first add your unfiltered object outlines (in pink to be parallel to this, or whatever color you choose) then second add your filtered object outlines (in green or some second color), you’ll end up with objects that failed the filter in pink and passed in green, which will give you essentially what you want.


#10

OK, good idea! I’ve tried it and it does a similar job indeed, although the Log normalization would be much better (left image) - with RescaleIntensity I either get a too dim image to properly see the neurites (middle image), or an image with clear neurites but just white pixels (right image). But it kind of works, thanks!