Cytomine Research @ Montefiore Institute, ULiège
Cytomine ULiege logo

(updated on Jan, 28th, 2021)


Overview

At the Montefiore Institute (University of Liège, Belgium), we are computer science researchers developing machine/deep learning algorithms and big data software modules to make life easier for multidisciplinary teams who have to deal with very large imaging data by enabling collaboration through sharing of images, algorithms, and results over the web.

The Cytomine, open-source, software (with a permissive licence) is a rich internet application using modern web and container technologies, databases, and machine/deep learning to foster active and distributed collaboration and ease large-scale image exploitation. The software can be used remotely, using a web browser, by life scientists to help them better evaluate drug treatments or understand biological processes using various imaging modalities (including whole-slide tissue images), by pathologists to share and ease their diagnosis, by teachers and students for image-based training purposes (e.g. histology courses), and by computer scientists who want to apply their image recognition algorithms on large imaging dataset. Beyond biomedicine, the software can also being used in various other application domains that involve large image datasets such as digital collections, industrial quality control, ....

In our research team, we continously contribute to the Cytomine open-source project initiated by our research unit in 2010. The official, base, version is validated and maintained by the Cytomine cooperative company. Our Cytomine ULiège version is forked from the official version and includes base features plus experimental features driven by our research projects and collaborations, but these are not yet fully validated.

Our main research interests and contributions are:

  • Development of efficient workflows and novel algorithms (tree-based machine learning and deep learning) for object detection, recognition, and segmentation in very large images
  • Development of novel algorithms and web user interfaces for multimodal imaging data
  • Benchmarking of algorithms on realistic datasets
  • Development of algorithms for user behavior analytics
  • Applications in the biomedical domain with a specific focus on digital pathology
  • Applications in any other domain with large image datasets (geology, digital collections, astronomy, industrial control, ...).
  • Software architecture, distributed software deployment, reproducibility, open science, ...

    • Ongoing research & results

    Here is an overview (PDF) of our research results, and illustrations below. In addition to our open-access scientific publications, our latest results are distributed through the ULiège's Cytomine open-source software repository (and later through the official repository managed by the cooperative).

    • Image segmentation techniques for the quantification of whole tissue slides.
      See e.g. Marée et al. ISBI 2014.

    • Workflows for sorting various types of cells, e.g. detect abnormal cells for early cytological diagnosis.
      See e.g. Delga et al., Acta Cytologica 2014; Mormont et al., 2016.

    • Cell counting algorithms in specific regions of interests within tissues.
      See e.g. Rubens et al. (in preparation).

    • Image classification and object recognition algorithms for diagnostic of for phenotyping, e.g. in developmental and toxicological studies.
      See e.g. Marée et al., PRL 2016; ISBI 2016 ; Jeanray et al., PLOSOne 2015; Mormont et al. CVPRW 2018 & IEEE JBHI 2020.

    • Anatomical landmark detection algorithms for morphometric change measurements e.g. in developmental and toxicological studies.
      See e.g. Vandaele et al. Nature Scientific Reports, 2018.

    • Tools for user behavior analytics e.g. in educational settings (Vanhee et al., Journal of Pathology Informatics, 2019).

    • User interfaces for multimodal datasets e.g. MS/multispectral imaging (Rubens et al., Proteomics Clin Appl, 2019).

    • Reproducibility and interoperability for deployment and benchmarking of image analysis workflows (Rubens et al., Cell Patterns, 2020).

    • Applications in various domains (not only histology) (see publications).

    News ( for more regular news)

    (January 2021-2027) We are involved in the BigPicture EU IMI project to establish the biggest database of pathology images to accelerate the development of artificial intelligence in medicine.

    (October 2018-...) We are involved in the COMULIS COST network for correlative microscopy.

    (October 2017-...) The NEUBIAS COST network WG5 choose the Cytomine software to develop its benchmarking platform (BIAFLOWS).

    (1 oct. 2017) Ulysse joined our research team @ ULiège. He is now our core developer working on extensions of Cytomine for 16bits/32bits/etc. images and machine/deep learning algorithms.

    (15 Sept. 2017) Grégoire, Christopher, and Renaud are now working for Cytomine.coop, a social cooperative to manage the open-source community and to provide services on top of Cytomine open-source software. Learn more about their services and their open company philosophy on Cytomine.coop.

    (Sept. 2017) The first worldwide massive online course (MOOC) in french on histology is over. It was a great success ! It will come back next year. It was organized by University of Liege and the FUN platform. More info, course teaser, and registration.

    (May 2016) Thanks to Regional-IT, Daily Science, and Le 15e jour for press coverage of our project and open-source release party.

    (Jan. 2016) Our Cytomine paper is formally accepted for publication in Bioinformatics journal.
    Cytomine is now released under an open-source software license and available on Github (see instructions below).
    You can still request an account on our demo server.

    (Sept. 2014) A vulgarization of some our research is available on University of Liege Reflexions website

    (Oct. 2010) Cytomine kick-off.

    Publications

    The main Cytomine publication is (Marée et al., Bioinformatics 2016). We list here our related publications.

    Software and source code

    We release our research results mainly through the R&D Cytomine open-source software ULiège repository, while the official, base version is available on the official Cytomine repository.

    To install and use the ULiège R&D version you can follow experimental version installation instructions and its full documentation wiki (it includes guides for administrators, developers, and a Cytomine guide for end-users). To install the official, base version, please follow official installation instructions

    Cytomine can be installed on large servers for large-scale studies but also on laptops for small-scale works (then loosing collaboration functionalities). You can also check the Cytomine user guide to get access information on our R&D demo server using default accounts. Access to an official, stable, version can be requested to Cytomine cooperative.

    We kindly ask you to cite Cytomine website url (www.cytomine.org) and our main publication (Marée et al., Bioinformatics 2016) when you use our software in your own work.

    Research team members and collaborators

    Raphaël Marée initiated the Cytomine research project in 2010. He is a senior researcher in machine learning/big data/computer vision and supervises the Cytomine big data research and development activities at ULiège.

    Ulysse Rubens is a software developer (Cytomine core modules and machine learning).

    Romain Mormont (PhD Student) is a researcher in machine learning (deep learning, transfer learning).

    Navdeep Kumar (PhD student) is a researcher in machine learning (deep learning, domain adaptation for multimodal images).

    Prof. Pierre Geurts and Prof. Louis Wehenkel are long-term machine learning collaborators.


    Previous contributors (team members or students) include: Loic Rollus (2010-2015), Benjamin Stevens (2010-2014), Renaud Hoyoux (2014-2017, now at Cytomine cooperative), Gilles Louppe (2013), Jean-Michel Begon (2014), Pierre Ansen, Julien Confetti, Olivier Caubo, Thomas Vessiere, Laurent Vanosmael (2017), Gino Michiels (2018), Laurent Vanhee (2018), Mehdy Ouras (2018), Guillaume Vissers (2019), Loïc Sacré (2019), Rémy Vandaele (2014-2019),...


    Through various collaborations (including the NEUBIAS COST Action), other contributors were/are coming from IRB Barcelona, Pasteur Institute (Paris), VIB Ghent, MRI Biocampus Montpellier, ...

    Developments and services related to the base, official, version are lead by the Cytomine SCRLFS team.

    Contact

    We recommend to use the Image.sc forum as the discussion channel for user related questions.

    For bugs or feature requests we recommend to post your issue on our Github repository.

    For any questions related to our research activities or for potential research collaborations or internships @ ULiège, feel free to contact us by e-mail: Raphaël Marée.

    To ask for a specific demo account, to support the open-source project, or for service requests (e.g. installation, hosting, maintenance, trainings, specific software developments, slide scanning), please contact Cytomine cooperative.

    Finally, if you want to visit us, we are based here:
    Quartier Polytech 1, 10, Allée de la découverte
    Montefiore Institute (B28)
    Tel. : +32 4 366 26 44

    Funding

    We are/were involved in the following projects:

  • BigPicture (2021-2027) funded by EU Innovative Medecines Initiative.
    This project will establish the biggest database of pathology images to accelerate the development of artificial intelligence in medicine.
  • BioMedaqu (2018-2022) funded by H2020 MSCA ITN network.
    This project aims at developing new image analysis modules for Zebrafish imaging.
  • DeepSport (2017-2021) funded by Wallonia DGO6.
    This project aims at developing new visualization and annotation modules for video data.
  • COMULIS (2018-2022) funded by EU COST Action.
    This project aims at developing new multimodal imaging visualization and annotation modules.
  • IDEES (Fondations Technologiques, 2017-2020) funded by FEDER (ERDF/Wallonia).
    This project aims at developing new big data software modules. Développement et déploiement web et distribué d'algorithmes d'analyse de données et d'intelligence artificielle pour les Big Data.
  • NEUBIAS (2016-2020) funded by EU COST Action.
    This project aims at developing image analysis reproducible benchmarking modules.
  • ADRIC (2017-2021) funded by Pole Mecatech, Wallonia/DGO6.
    This projet aims at developing new AI modules for industrial quality control.
  • HISTOWEB (2014-2017) research grant n°1318185 funded by Wallonia/DGO6.
  • SMASH (2012-2014) research grant n°1217606 funded by Wallonia/DGO6.
  • CYTOMINE (2010-2016) research grant n°1017072 funded by Wallonia/DGO6.
             DGO6          COST          Feder enmieux          Pôle Mecatech          Pôle Mecatech          EU Marie Curie Innovative Training Networks IMI University of Liege