How to quantify intensity of fluorescent of mycelium

Image Analysis
#1

Hi all,

I’m a new user of ImageJ, and I’m interested in measuring the fluorescence intensity of the mycelium in a transformed isolated (GFP) which is infecting plant tissue. I’ve read a lot of ImageJ articles, but I’m a little confused about how to do that. I was trying to adjust the threshold in order to obtain the measure but my results vary a lot, so, I read an article about segmentation and I tried to use statistical region merging and robust automatic threshold selection but I’m not sure if this could work for my pictures because the results are very different between them so I would like to know which is better to eliminate any bias in thresholding or please tell me if you have another suggestion

I’ve included a sample image below of the fluorescence I’m hoping to measure. I’d like to measure the fluorescence intensity of the mycelium (green)

Glenlea_SGFP-K1-2NC-4_3dpi_ 4x BF_ColorComposite.tif (11.0 MB)

Thank you in advance for your assistance!

#2

  1. It looks like there are some severe scanning or imaging artifacts in the horizontal direction which will cause problems for any kind of measurement or thresholding.

  2. Thresholds for measuring intensity of a fluorophore should generally not be performed using that fluorescent channel. Measuring fluorescence intensity - #2 by etadobson

  3. If you have a lot of sample variation, rather than using a fixed threshold, you may be able to use a standard method, like Otsu or one of the others (listed here Auto Threshold - ImageJ). This can be very dangerous to do blindly, though, and you should doublecheck all of your results as samples like negative controls can end up with VERY different thresholds due to lack of signal.

2 Likes
#3

Hi,
Thank you for the reply, I followed your advice and I used Auto Threshold (Otsu). However, there is a lot of variation between my samples, and I do not trust the reliability of my results. Maybe Do you have another suggestion to improve the reliability?

Note: I’m using 16 bit images (GFP filter)

Thanks!

#4

My only recommendation would be to look into “some kind of normalization.”

How exactly that would work would depend on the source of the variation, whether it is within images or between images, and whether there are batch effects between groups of images. I haven’t dealt with it much, but there should be other forum posts on normalization or batch normalization.

#5

I will do that, Thank you!

1 Like