Closed mkeays closed 4 years ago
Hi Maria,
I don't believe any IMC images were used to train the CellProfiler nuclei detection U-Net that Cytokit uses so it's probably not a great idea, or at least it would be hard to justify applying it without validating the model for that imaging modality first.
Out of curiosity though I ran it on one of your H3 images and the results are better than I thought, but still not as good as fluorescent images (w.r.t. nuclei detection):
I think you'll probably get more mileage out of some package or plugin specifically for IMC. Sorry!
Hi Eric,
Thanks for your quick reply and for explaining -- I will look at some other packages.
Thanks! Maria
Hello,
I am looking for open source, automation-friendly software to do end-to-end processing for imaging mass cytometry data, and I was wondering if Cytokit would fit the bill? I have data similar to the example dataset here.
Thanks, Maria