search

BodenmillerGroup / ImcSegmentationPipeline

A pixel classification based multiplexed image segmentation pipeline
https://bodenmillergroup.github.io/ImcSegmentationPipeline/
MIT License
83 stars 35 forks source link

ilastik classification #76

Closed sailseem closed 2 years ago

sailseem commented 3 years ago

Dear developer, I am wondering the classification of ilastik in the example pannel csv files, which antibody should be chose as 0, and which should be 1. It's kind of confused when applying to our own data. Thanks a lot. image

nilseling commented 3 years ago

Hi @sailseem

the 1 in the full column means that these channels will be analysed (e.g. written out as .tiff files and feature extracted). The 1 in the ilastik column means that these channels will be used for ilastik pixel classification. So please select the channels from which you think they will be useful to identify the nucleus, cytoplasm, and background.

Also have a look at the documentation here: https://github.com/BodenmillerGroup/ImcSegmentationPipeline/blob/main/documentation/imcsegmentationpipeline_documentation.pdf

Or the online docs (under development): https://bodenmillergroup.github.io/ImcSegmentationPipeline/prepro.html

nilseling commented 2 years ago

Closing this under the assumption that the instructions resolved the issue. Please reopen if anything regarding the panel entries is unclear.