Hi Eitan,
it looks like a problem that could be solved by the Euclidian Distance Map
(EDM):
- Create a binary image of the cell, invert it and run
Process>Binary>Distance Map.
- Find the values of the distance map at the positions of the bacteria. An
automatic way of doing this is the following:
Create an image suitable for finding the bacteria with Process>Binary>Find
Maxima, create a "Single Point" output and use the 'AND' operation of
Process>Image Calculator to combine the result with the EDM. Now, you will
have a point for each bacterium with a value corresponding to the
distance.
You can use, e.g., the particle analyzer to make a Results Table out of
these.
If you have selections, not images as an input, you can convert those to
images by Edit>Selection>Create mask.
Note that the EDM in ImageJ only works for distances up to 255 pixels.
Hope this helps a bit,
Michael
____________________________________________________________________
On Sat, December 20, 2008 23:11, Eitan Z. wrote:
> Hello to the members of the forum,
>
> I have several series of time-lapse microscopy images of bacteria colony
> attachment to cells, what I'm trying to measure is their changing
> distance
> to cell boundary.
>
> Since the cells themselves tend to have some movement of their own in
> the
> field, what I've done is:
> 1) I've manually drawn freehand line selection for every time-point,
> 2) Drawn a freehand circular around ononebacteria colony in each time
> point.
> 3) Measured the centroid and center of mass of the circular ROIs (which
> are close, since the intensity of the colony is quite evenly
> distributed).
>
> So now I have a series of corresponding pairs of freehand line ROIs and
> quasi-circles and I want to measure the distance between the particles
> center and their appropriate lines.
>
> I'm pretty new to ImageJ, but I figured someone probably done this
> before,
> but I couldn't find a plugin or how-to for this challenge...
>
> I also looked for a freehand line to point distance solution, but
> couldn't
> find one either.
>
> Can anyone help?
>
> Thanks,
> Eitan.
Locked
