OmeSliCC

Ome(ro) Slide Image Conversion and Compression pipeline

OmeSliCC is designed to convert slides from common formats, to optimal OME formats for deep learning.

This includes converting from Omero and extracting metadata as label information.

For support and discussion, please use the Image.sc forum and post to the forum with the tag 'OmeSliCC'.

Main features

• Import WSI files: Omero, Ome.Tiff, Tiff, Zarr *, Ome.Zarr/NGFF *, common slide formats, common image formats
• Export images: Tiff, Ome.Tiff, Zarr *, Ome.Zarr *, common image formats, thumbnails
• Zarr image compression (lossless/lossy)
• Image scaling using target pixel size
• Omero credentials helper

*Zarr currently partially implemented. Also see OME NGFF

Running OmeSliCC

OmeSliCC is 100% Python and can be run as follows:

• On a local environment using requirements.txt
• With conda environment using the conda yaml file
• As Docker container

Quickstart

To start the conversion pipeline:

python run.py --params path/to/params.yml

See params.yml for an example parameter file. The main sections are:

• input: providing either a file/folder path, or Omero URL
• output: specifying the location and desired format of the output
• actions: which actions to perform:
  • info: show input file information
  • thumbnail: extract image thumbnail
  • convert: convert to desired image output

To encode credentials for Omero access:

python encode_omero_credentials.py --params path/to/params.yml

To extract Omero label metadata to text file:

python extract_omero_labels.py --params path/to/params.yml

Documentation

https://franciscrickinstitute.github.io/OmeSliCC/

Acknowledgements

The Open Microscopy Environment (OME) project

The Francis Crick Institute

• The Software Engineering and Artificial Intelligence team
• The Turajlic lab