How to quantify depigmentation in flatworms

Tags: #<Tag:0x00007fa30b9f9808> #<Tag:0x00007fa30b9f9560>


Hello community,

We are two undergraduates who are new to imagej and are looking to quantify the amount of pigment in small flatworms, planarians (similar to c elegans), and are wondering what the proper way to use imagej to quantify that is. Attached are images of worms before and after depigmentation, and we were hoping someone could help us make a program to quantify the differences between depigmented worms. To do so we must find the pigment spots which are darker to their local background. A typical threshold finder isn’t working because the lighting differs between the worms and sometimes even in one worm photo on its own. Image of example worms


Do you really? I’m not saying this is not the way to go but if you’re looking for an overall change from dark (pigmented) to light (depigmented) have you thought about just measuring the mean brightness?

Below, I’ve converted the image from RGB to a hue, saturation and brightness (HSB) stack using [ Image > Type > HSB Stack ] and (after selecting the worms with a magic wand tool - tolerance ~130) measured the mean brightness per worm.


ID Area Mean StDev Mode Min Max
1 96629 119.328 23.723 127 14 177
2 84852 137.058 24.412 139 23 202
3 92169 136.037 29.134 143 16 204
4 75323 127.198 28.808 144 18 171
5 82193 126.749 32.397 150 18 176
6 76609 158.385 33.239 177 39 200
7 79315 158.956 30.993 189 19 202
8 70548 147.907 34.074 176 29 194

(Remember that this is the mean brightness not the mean intensity!)

You mention some variability in your samples which you would need to address, but I just wanted to suggest trying something simple to begin with.

You may also want to think about normalising your values to a fully pigmented sample (or depigmented sample) that way you always have an internal reference in case of different imaging parameters.


Thanks a lot, your input helped give us a head start on how to use the program.