Counting OVERLAPPING cells (Watershed not accurate)

[anne]

Hi!

I am trying to analyze this image using ImageJ. I want ImageJ to count the stained cells. However, due to overlap of multiple cells, when I make the image binary I lose the original boundaries and hence, ImageJ is unable to watershed correctly. Also, seems watershed only works well with circular objects.

I used Process --> Find edges which outlines the cells relatively more accurately. To analyze particles, I would have to make image binary or threshold and in both circumstances the cluster of cells is counted as 1 cell by 'Analyze Particles'.

Any advice on how to count cells would be greatly appreciated!
Thanks!
A

[jerie]

Hi Anne,

it may be possible to approximate the count by estimating the average size
of one nucleus under this conditions, then divide the entire thresholded
area by this average size.

Certainly there is a lot of room to improve the imaging, so depending on
the aim of the study it might be more efficient to repeat the experiment
seeding less cells.

Hope that helps,
Jens


Dr. Jens Rietdorf, visiting scientist @ center for technological
development in health CDTS, Oswaldo Cruz Foundation Fiocruz, Rio de Janeiro
Brasil.

On Sat, Jul 15, 2017 at 3:53 PM, anne <[hidden email]> wrote:

> <http://imagej.1557.x6.nabble.com/file/n5019069/20133566_
> 1447724931939733_2051726349_o.jpg>
> <http://imagej.1557.x6.nabble.com/file/n5019069/20133594_
> 1447725395273020_1132230020_o.jpg>
>
> Hi!
>
> I am trying to analyze this image using ImageJ. I want ImageJ to count the
> stained cells. However, due to overlap of multiple cells, when I make the
> image binary I lose the original boundaries and hence, ImageJ is unable to
> watershed correctly. Also, seems watershed only works well with circular
> objects.
>
> I used Process --> Find edges which outlines the cells relatively more
> accurately. To analyze particles, I would have to make image binary or
> threshold and in both circumstances the cluster of cells is counted as 1
> cell by 'Analyze Particles'.
>
> Any advice on how to count cells would be greatly appreciated!
> Thanks!
> A
>
>
>
> --
> View this message in context: http://imagej.1557.x6.nabble.
> com/Counting-cells-ImageJ-tp5019069.html
> Sent from the ImageJ mailing list archive at Nabble.com.
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html