I'm working with images of cells stained for beta-galactosidase but the blue signal is not very prominent compared to the amount of gray/background signal.
1. what methods do you use for quantifying blue signal from a phase contrast image?
2. are there ways to filter the image so that i can eliminate some of the background without affecting the blue signal?
Thank you!
Sindhu,
You can try analyzing your image with ImageJ software. This software has option where you can split the image into different channels. Once you have the blue channel image separate, you can quantify by deducting background signal.
Good luck
Hi Sindhushree,
ImageJ's channel split option is not optimal if you are working with a phase image with a histological XGal stain. The blue color is not a 'clean' blue like in a false colored fluorescence image. Since, the XGal stain contains a lot of green and red components, if you filter only for the blue part you may end up worsening your signal to noise ratio (see in the image).
I have found that working with CellProfiler (see link below) and using their UnmixColors module works well for XGal stainings.
Best,
Anton
http://www.cellprofiler.org/
Hi Anton,
Thank you for your response - once you retrieve the gray-scale image from UnmixColors, do you know of any modules on CellProfiler, ImageJ, or MATLAB that might be useful in quantifying the number of x-gal positive cells or scoring the amount of blue signal coming from x-gal positive cells?
Thanks again for your help - I really appreciate it!
~Sindhu
Hi Rajesh,
Thank you for your suggestion - can you recommend a specific ImageJ protocol or tutorial for quantifying amount of blue signal present within the boundaries of cells in a phase contrast image?
Thank you - I appreciate your help!
~Sindhu
Hi Sindhu,
that depends on how your images look like. If your cells are spread out and differ highly from the background or have clear borders you can attempt an object recognition module.
Frankly, often it is faster to count manually, but if you can upload an example of your work, maybe I can come up with different ideas.
Best,
Anton
Hi Anton,
I've attached a couple of sample images:
SampleImage_RG_question2: before rolling ball background subtraction on ImageJ/raw image
SampleImage_RG_question3: after rolling ball background subtraction on ImageJ
The reason I'm trying to avoid counting manually is because I'm not sure if we have a sufficient number of fields from each sample to approximate the percentage of x-gal positive cells. The cells were imaged while in culture and not stained and preserved on slides as is typical for histology samples.
Thank you for your feedback - I very much appreciate it!
~Sindhu
Hi Sindhushree,
were you able to figure out how to analyze images for X-gal staining?
I am researcher at McGill University trying X-gal. My cells tends to grow in cluster and dificult to count manually.
Also, I have different shades of blue and I am wondering which color should I consider as true and not false positive.
Please help me guys.
Regards,
Shraddha
Fluorescein based β-galactosidase assays in GFP channel such as the "CellEvent Senescence Green Detection Kit" offer a better alternative to traditional X-gal based assays, and it's easier to analyse the images. You can use CellProfiler with "MeasureObjectIntensity" workflow and set a threshold for classifying positives and negatives based on intensity of individual cells.
- Badges
- Science method
- Similar topics
- Biological Science
- Histology
- Staining