Quantifying Stained Liver Tissue Area Using ImageJ

Measuring stained area of a fibrotic liver section.
Masson's trichrome stain

[@georgsemmler8645]

Thank you for the great video. However, is it possible to measure the "fibrotic" area without also including the area of the nucleoli? Which is the case in your example! Thank you for the help!

[@vongola113]

thabkyou very much for the clear instructions it was very helpful

[@sonntana]

This is method is user friendly and simple, thank you for the video. Can you explain a little bit more, is there any good way to express our data in to graph? Which measurement/value should we use? Thank you.

[@edwardmoreirabahnson803]

I think the video is great. However, I'm confused about something and I'd appreciate some clarification. You're selecting your threshold with the red channel. Despite that, you actually select what's not red (blue?). When I tried to do this as you described it, the selection were all the " light" pixels of that channel or put differently all the non fibrotic areas. I had to use the blue channel (which has the fibrotic areas as " light pixels" ). Could you comment on this? I thank you in advance.

[@ChengIan]

You could use the "dark background" as shown in 0:51 to switch between highlighting pixels of interest and background in one channel. The video illustrates how to take measurement using ImageJ. Conditions such as which channel to take measurement from is up to you since images can be different.

[@JoshuaAdamsMDPHD19]

Very helpful; just what I needed to know! Thanks!

[@channelutb3657]

How to download image j?? Iam a medical student, now i needs this application to help my reasearch about the diabetic nephropathy

[@notneedit]

@@channelutb3657 Did you find the way to use quantify fibrosis! I just did a study on the same topic, but it really gave me a headache!

[@gazelephant]

@cheng Ian, thanks a lot for the explanation.

[@koolsool14]

informative video Cheng. I would like to know can we use this method for both cytoplasmic dab staining and nuclear dab staining??

[@valeriacapaci33]

Great method. is there any way to cite it on a publication? thanks

[@oralfixij]

This video was extremely helpful. One question - how do you decide where to set the thresholds? Isn't it somewhat arbitrary? If you're comparing between two conditions, making measurements of multiple images, how can you do so accurately? Thanks!

[@ChengIan]

You need to set up a standard first, ideally you don't create your own but adopt something that is agreed upon and verified by the scientific community. If you are working on something entirely new, at least you have to be consistent in setting thresholds for all of your images (e.g. using the same range for all images), and I think in that case your biggest concern is to justify your decision.

[@ivancarranzam.8208]

Thank for the video. !!! It´s perfect for my samples !!!!

[@IMuril0]

what does the "limit to threshold" options actually does?

[@gazelephant]

Great tutorial. I just have one question: the area you get at the end is measured in pixels, how would we go about putting that into, let's say, micrometers? Thanks.

[@ChengIan]

when you took the images, you should have magnification and field of view, the information you can use to convert from pixel to length scale let's say you have an image 800x400 in pixels and the field of view is 2 mm x 1 mm, then each pixel is 2.5 micrometers

[@giantsockmonkey]

Hi, very helpful video. Why do you need to make a montage of the stacks?

[@ChengIan]

giantsockmonkey it's just for show, not useful here

[@ChristopherNicholasReeves]

Hey! Thanks for the helpful video! I tried it out on one of my images, and I THINK it worked, but is there any chance I could send you the image and see what you get? As I'm just not sure about my analysis and skills with ImageJ.

[@ryanmcallister3944]

Is there a publication to cite this method?

[@hollier8744]

This supposes the area of the image occupies the entire space in the viewing field. What do you do if you have a tissue microarray slide image, for example, that does not (i.e. has outer boundaries)? Is there a way to segment that to determine total area of the tissue? Hand tracing isn't appropriate as there is no single boundary like a cell membrane. Thanks!

[@ChengIan]

Hollie Ryan I think you could use thresholding to get the area of all the microarray cells, since there should be a clear boundary, then threshold again to obtain the area of the tissues inside all the cells.

[@hollier8744]

I ended up doing that and getting decent results (so far!). Thanks for your video and reply!

[@ChengIan]

Glad to be of help:)

[@snehaswapnil7485]

Very helpful video. Thankyou so much :)

[@ChengIan]

Glad it was helpful!

[@jayapradha7448]

Hlo sir, I saw your video today only. really It was a wonderful video. So I have one query which thing I have to take for result, I mean for make the final graph

[@ChengIan]

The area and %Area are your results, depending on which one you would like to report

[@Linda-hd5ew]

Thank you so much!

[@dalaydayjohnlloyd5532]

how do you get this kind of image from stained fabric?

[@ChengIan]

this dye is specific to staining cells, I don't think it works on fabric, there should be specific dyes for that. Then the analysis should be very similar

[@dalaydayjohnlloyd5532]

@@ChengIan thank you very much! I'll look more into that

[@vradojc1]

Is there a way (already a part of ImageJ), a plugin, or a macro script for batch processing of slides this way? Completely new to ImageJ and I cannot find one. I guess batch processing for the first step (changing type to RGB stack) would be tremendously helpful as well.

[@ChengIan]

Vedran Radojcic there is a way to create your own script, go to plugins → macros → record… and then just perform any sequence of operations as you would, imagej will record your operations and compile a script for you to run automation in the future.

[@vradojc1]

Thanks for the reply. Since this records the file name in the first line, it will run macro only on that file if open. Do you know how to name file in that command line to apply to all open images?

[@ChengIan]

you will need to modify the script to run through a sequence of images automatically, i'm not very well versed in imagej's micro language, but you could google "imagej macro open image sequence" it should lead you to webpages with detailed discussions on how to do it

[@vradojc1]

Thanks, sorted batch RGB stacking and output analysis.

[@ChengIan]

Have you tried opening all the images inside the folder and then run measurement on all of them so that all results can be grouped in one window?

[@richelley7]

Would this also work for quantification of other stains such as PAS stain or is it specific for Masson's trichrome stain?

[@ChengIan]

Chelley i think the thresholding method should work on other stains as well

[@richelley7]

Thank you. Are you able to suggest where I may possibly find reference for what threshold values I should use?

[@ChengIan]

Usually it's by your own judgement, but keep the threshold values consistent across all images. There might be published protocols, but very often authors skip over the details on data analysis. I think asking those authors could be a good idea to get the values they used.

[@richelley7]

Cheng Ian Thank you

[@dinidamayanti314]

I am interesting about your video.Can I use this software to quantification of immunohystochemestry slide . If yes, how can I get this software? Thankyou

[@ChengIan]

Dini Damayanti It's called ImageJ, you can get it from: imagej.nih.gov/ij/

[@dinidamayanti314]

Thankyou for your information

[@ChengIan]

you're welcome!

[@EB-hp9le]

Hi, I’m in the middle of writing up my dissertation for my masters. As part of my dissertation, I have to perform an image analysis of fibrotic lung tissue using image J and I find myself being slightly stuck. Would you be willing to have a quick chat sometime?

[@ChengIan]

Sure, what's your email to connect?

[@EB-hp9le]

@@ChengIan Hi, I think KZbin deletes my comments stating my email address, can you please confirm if you received it?

[@ChengIan]

@@EB-hp9le No I didn't, try my gmail, user name: amos.eclipse

[@dcdcjr3]

does this work with h and e staining?

[@ChengIan]

i think it does, try picking the right color frame and adjusting the threshold

[@tcs3600]

This method works best if your colors are "pure" red or blue. For H&E and other stains, try using the Color threshold (select by hue, not limited to R/G/B) or the Colour Deconvolution plugin by Gabriel Landini.

[@Anexoticadventure]

I was seeking an answer to this exact question - appreciate the feedback! I might have a follow-up... :-)

[@passionofJC]

thanks a million..:-)

[@thurayae]

thx alo0o0ot....