about cell segmentation on phase-contrast cell image

[cube...@gmail.com]

Hi.

I am making an attempt at cell segmentation on phase-contrast cell image. Purpose of cell segmentation is quantifying fluorescent signal per a single cell. 

Quantifying fluorescent signal can be done somehow using 'spot detector' and  'ROI measure' of ICY, but precise cell segmentation remains unsolved.

Actually, there are a few options for cell segmentation. First, with cytosol staining dye, I can stain whole cell, but I have no available color channel. Second, making ROI by drawing polygons one by one on fluorescent or phase-contrast cell images is also possible option, but this is not efficient as well as time-consuming.

So I decided to do cell segmentation and make ROI (for qualifying fluorescent signal) on phase-contrast image.

I would be really appreciated it if I find how.

Please give me some hints or advices. Thanks in advance.

Best regards,

Junsung, LEE.

[Alexandre Dufour]

Hi,

There are a few solutions in Icy that you could use to automate the cell segmentation process, although it depends of the modality and quality of the images. You could try the Active contours plugin (http://icy.bioimageanalysis.org/plugin/Active_Contours) or the Active Cells plugin (http://icy.bioimageanalysis.org/plugin/Active_Cells), or simpler tools such as the HK-Means (http://icy.bioimageanalysis.org/plugin/HK-Means) for fluorescence data.

In the meantime, would you mind sending one or two sample images? We could try to figure what works best and send you the corresponding protocol...

Alexandre

--
You received this message because you are subscribed to the Google Groups "Icy imaging" group.
To unsubscribe from this group and stop receiving emails from it, send an email to icy-software...@googlegroups.com.
To post to this group, send email to icy-so...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msg/icy-software/-/ldL0MnHc0FYJ.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[Paul, Hsieh-Fu Tsai]

Hi Guys,

I'm Paul working with alot of cell migration studies with time lapse microscopy (primarily phase contrast).

I have been looking for a software that can help me segment through the time lapse data accurately. But I'm not so good at image processing so i haven't locate one that works for me.

I wonder if it's possible for Icy to do these data?

one is the raw data where background is very visible, and the other is one i tried to do some processing.

I really hope i can find a solution to segment the cells into ROIs that correctly recognize the entire cell body including lamellipodia.

If anyone knows which direction works best, I really appreciate if you can point me in the right direction.

Best,

Paul

[Alexandre Dufour]

Hi Paul,

Unfortunately your problem is one of the most complicated ones around. Not only is this tricky to find the location of cells accurately in order to properly track them over time, but the precise extraction of the cell boundary is very critical in such modality, notably since the contrast close to the membrane disappears more often than not and blends in with the background.

Now because I don’t want to leave you with only negative comments, here are some ideas to think about:

 — I have seen some recent papers on cell segmentation using deep learning that have managed to extract the cell boundary with some success, although I am sure these methods have their own caveat (see for instance this: http://arxiv.org/pdf/1505.04597v1.pdf)

— You seem to be working with NIH3T3 cells, and depending on the duration of your time-lapse movies, I would suggest you look towards the fluorescent version of the sane cell line, because once in fluorescence the segmentation and tracking problems are considerably easier.

Sorry for not having a ready-made solution, but I do hope these are interesting pointers.

Alexandre

To view this discussion on the web visit https://groups.google.com/d/msgid/icy-software/cadfc71a-3e15-4ad0-b6cd-449c8a5fcdf3%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
<Result of 20160207_nih3t3_chip_300mvmm002_xy08-20002.tif><20160207_nih3t3_chip_300mvmm002_xy08-10002.tif>