3D project in other modes (eg. StDev)?

I frequently find that Z-projections of my stacks are most informative when viewed as Standard Deviation projections. (Incidentally, I’m not entirely clear on what this projection mode is doing; is it showing the values that are more than one St.Dev from the mean of the population of values?). Unfortunately, this method of projecting a stack is not available in the 3D projection menu (just “nearest” “brightest” or “mean”, where brightest seems to be doing the same thing as a maximum projection in Z).

Has anyone written a macro that can do St.Dev projections in 3D? Would anyone be interested in this?

Hi @bryanc,
We implemented such a feature in KNIME Image Processing.
The Projector Node allows you to project Images with the following five options:

  • Max
  • Median
  • Average
  • Min
  • Standard Deviation

Take a look at our forum post on how to get started with KNIME Image Processing and its ImageJ integration.

2 Likes

Thanks,

I downloaded KNIME and will try to experiment with it when I have time. It looks like it will have a steep learning curve.

I use Fiji both for my own research, and for teaching undergraduates basic microscopy and image processing; it seems to me the KNIME will be more than they can handle (Fiji is very intimidating for many of them). So my preference would be to do this within Fiji/imageJ if possible.

Hi @bryanc,

maybe the material from the ETH Zurich (https://github.com/kmader/Quantitative-Big-Imaging-2015) helps you getting started. They used KNIME for their entire BioImaging lecture. More material (Videos, Example Workflows and more lectures/talks) can be found on http://knime.imagej.net. Maybe this helps you getting started. Let us know if you face any further problems or need any further help.

Btw: The projection issue should also be solvable using the ImageJ Script Editor (http://imagej.net/Using_the_Script_Editor) in combination with ImageJ-Ops (for example see https://github.com/imagej/imagej-tutorials/tree/master/using-ops and https://github.com/imagej/imagej-ops).

Cheers,

Christian