I was wondering if you could provide some insight into whether V3D can or could potentially accomplish our goals for 3D segmentation. We are acquiring confocal images of various bacterial communities (12 bit, 1024x1024, 20 um z-stack sampled @ Nyquist ). These images often have between 8 and 16 different color channels. Ideally, we would like to perform local thresholding algorithms to individually segment each channel in 3D. Our primary purpose in segmenting each channel is to determine the 3D centroid position of each bacterium within each channel. Any information you can provide concerning V3D's capabilities in relation to our project would be extremely helpful. Thank you in advance for your time.

Yes, V3D (or V3D plugins) can do the segmentation. Previously V3D came with a 3D cell segmentation module (done by Fuhui Long) with which you can specify which color channel you want to use to segment. Now we are in the process of porting such a module to a V3D plugin. We have not released it yet.

But any one can use their segmentation methods and use it with V3D, under the plugin mechanism. The V3D main program just needs to pass the color channel information to the plugin and can also display the segmentation results in 3D.

best,
-hanchuan

Originally posted by :

Is it possible to do this segmentation with up to 16 different color channels in one image? Thus far, I have only seen 3 channel RGB images displayed in V3D.

v3d currently only displays the first 3 color channels (and map them as RGB). If you want to see more channels, you can use the color blending function, or color channel flipping function, etc to see them.

One of course can easily write a simple plugin to pick which color channel should be visible at a time.

All the color channels will be read from a file; there is no limitation on how many color channels you can have. Therefore, sure, you can segment images with even 32 color channels.

I am working with different information of the same images stored as different image stacks. We would like to open them as different channels of the same image (multichannel image).
However, what I see now is that the first three channels are always mapped to RGB. Is there a possible way to preserve the original color space of the image stacks loaded as channels?

Also, I would like to open together, as a multichannel image, stacks that contain different number of channels between them. So far I could not load those all together (for example one dataset which is single-channel grayscale and one that is RGB). Could this be done?

I am thinking that a way to do that could be having some metafile that would specify the channels and their color space when loading them...?

Thank you,
-Mariona

If you have N channels in your image, you can load it first and use the right-bottom color mapping tools in the tri-view window to map different channels (you can select as many as you like) to different colors....

you can extract individual channels for segmentation, or combine multiple channels from different images (with the same size) to be one single image stack using tools from the main menu.

-Hanchuan




Originally posted by marionac: