Quantification of necrotic regions on tumor tissue slides by QuPath

[YurikoOgura]

Dear Qupath users and developers,

I am a complete beginner at both pathology and image analysis, so I’m very sorry for posting a very basic question.

I am now trying to quantify “necrotic regions” on H&E-stained tumor tissue slides deposited in TCGA.

(I want to evaluate the relationship between the degree of tumor tissue necrosis and their clinical outcomes, various molecular traits, and so on.)

My questions are ;

1. Is QuPath suitable for my experimental purpose?

2. Is QuPath capable of morphologically distinguish necrotic area from others (e.g. normal cancer cells and tumor stroma)?
   Is a process of training an image classifier required?

These may be silly questions, but I would appreciate it if you could give me some information or advice.

Sincerely,

[carcen (Caner Ercan)]

Yes, it is possible with QuPath. I would suggest you to give look at this method:

QuPath Guide: Creating classified annotation areas using SLICs

Quick tutorial on generating annotation level areas from SLICs. Image courtesy of La Jolla Institute of Immunology Microscopy Core shared image resource. Not going to go into a whole lot of details here as the entire method will likely be obsolete once the Pixel Classifier is fully functional. Generate your primary area with an annotation, either simple tissue detection, createSelectAllObject() or something else like drawing a square. Generate your SLICs. There isn’t any one right answer fo…

[Research_Associate (MicroscopyRA)]

There are a decent number of posts if you search for QuPath and necrotic/necrosis - you may get some ideas from them. One of the issues seems to be the rather wide variety of visible appearances of necrosis at different stages or for different regions. Some will be easier to identify than others.
For example this thread has a few different approaches, not all relating to QuPath.

Quantification and pattern recognition of necrosis

There are many different types of necrosis and appearances depending on the tissues where this happens. You would need to see features at the cytological level, not at low power, even if the texture is slightly different from the surrounding tissue.

[YurikoOgura]

I really appreciate your advice.
It seems very difficult for me, but I’ll do my best!

Thank you very much.

[YurikoOgura]

I really appreciate your advice.
I missed this topic, so it was very helpful for me.

Everything seems very difficult for me, but I’ll do my best!

Thank you very much.

[petebankhead (Pete)]

Hi @YurikoOgura I think it is a difficult problem. QuPath doesn’t contain any features specifically to quantify necrotic regions. But as @carcen and @Research_Associate describe, QuPath contains lots of image analysis and machine learning tools that can help solve that problem.

However, learning how to use those tools and figuring out how to combine them to create a custom algorithm for this purpose remains challenging for anyone – and how well it can be solved with QuPath alone often depends upon how variable your images are.

[YurikoOgura]

I really appreciate your kind advice.
I would like to consider the best way for my purpose by referring to various posts and comments in this community.

Thank you very much !!