Re: cytoseg output

[Rick Giuly]

Hi Kurt, glad to hear you're getting nice results!

Right now you're looking at the "composite" folder. That one is just for
visually checking how the segmentation went. The one you want to use now
is the "resized" folder. That is essentially just the output without the
background image.

Now, if you want to make the mitochondria into individual objects, you
can do that with IMOD or Fiji. I have instructions on the cytoseg page
for both under:

Using Cytoseg with IMOD (make training data, view results, and edit results)

Using Cytoseg with TrakEM2 (ImageJ/Fiji) (make training data, view
results, and edit results)

(Also, you could do a threshold, followed by 3D connected components in
Fiji, and then open that label results in Amira.)

Once that is done, Fiji can give a count. I believe that Fiji has
extensive tools for analysis. Also, IMOD can give a count. I'm not sure
how many analysis tools IMOD has.

Hope that helps, let me know if you have more questions

-Rick


kurt weiss wrote:
> If I open them with FIJI, I can see a slight color(gray)
> variation where the green used to be, but no green. When I open them in
> Amira, I can see green, but I don't have the usual 'label fields' option
> to let me threshold by color. And Ilastic won't open them at all - some
> error about "objects cannot be broadcast to a single shape". In any
> case, I wondering how the green is drawn on - is it a layer? And then if
> you have any recommended software/plugins for loading the data in 3D and
> actually connecting the mito to get counts and measurements of the total
> number of true mito in the stack.
>