This repository contains the supporting material for the technical paper on the pyramidal OME-TIFF format and the BIo-Formats implementation submitted for the 15th European Congress on Digital Pathology (ECDP).

OME-TIFF is an open file format allowing to store multidimensional images combining the OME imaging metadata model and the TIFF specification.

In February 2019, the OME-TIFF specification was updated to add backwards compatible support for multi-resolution images like Whole Slide Images. The sub-resolutions section of the OME-TIFF specification describes how downsampled images or sub-resolutions must be stored using the SubIFDs extension tag of the TIFF specification.

Public examples of WSI images with various modalities (brightfield, fluorescence, Z-stack, multichannel) are publicly available as OME-TIFF samples.

The Bio-Formats library supports the reading of over 150 proprietary file formats (PFFs) including multiple Whole Slide Imaging (WSI) formats. In addition, it acts as the reference Java implementation for reading and writing OME-TIFF files.

Bio-Formats 6.0.0 introduces support for the updated OME-TIFF specification with sub-resolutions. Binaries, API documentation and technical documentation of this release are all publicly and freely available.

The Bio-Formats bfconvert command line utility was used to convert WSI PFFs into open OME-TIFF with different compressions, tile sizes. When no pyramidal levels are present in the original image, this command allows to specify a number of sub-resolutions to be generated with a given downsampling factory.

To test the generation of multi-resolution OME-TIFF files, we used a set of original WSI PFFs covering various modalities (brightfield, fluorescence, multi-channel, multi-Z, floating point). The generate-samples.sh script contains the conversion parameters used for generating the derived OME-TIFF files.

It is possible to reproduce this conversion using the Docker image included in this code:

    $ docker build -t converter converter
    $ docker run --rm -it -v $(pwd)/data:/out converter

This will generate OME-TIFF files under the data folder. Alternatively, you can directly download the official OME-TIFF samples:

    $ wget -i ometiff_list.txt -P data/

The OME-TIFF datasets generated in the previous section have been validated using two open-source application using Bio-Formats for reading WSI:

• QuPath a software platform for whole slide image analysis
• OMERO a data management platform

A milestone release of QuPath 0.2.0 has been released including support for OME-TIFF pyramids. After downloading and extracting the application, images can be imported using File > Open.

A full OMERO application using OMERO 5.4.10, Bio-Formats 6.0.0 and OMERO.iviewer can be deployed via Docker compose. All the images under the data folder will be auto-imported at startup time:

$ docker-compose build
$ docker-compose up -d

After deployment, you can log in to http://localhost:4080/webclient using the root user with the password omero.