Machine Learning for Image Analysis



Up-front online sessions: October 2, 9 and 16, 2018
Workshop: October 29-31, 2018
Follow-up online sessions (optional): November 9 and 16, 2018



Modern electron, light and X-ray microscopes generate very large datasets that challenge our capacity to store, process and share them. The analysis of the dataset is the main bottleneck, manual analysis of a typical light-sheet or volume EM dataset can easily take many “person-years”. To meet these challenges the development of machine learning-based methods has been a topic of intense interest. Machine learning methods enable computers to learn without being explicitly programmed. For a computer to learn, it needs to have some initial data on how to do a specific task. The computer will find statistical patterns in the data that will enable it to establish the algorithm by which future data will be analysed.

“Facilitating the uptake of the latest machine learning methods in service facilities is of high importance, this course will help to spread best practice and train scientists in the skills needed to apply these methods” says Vera Matser, course organiser and CORBEL work-package leader for training.

The research of Matteo Rauzi, who runs a CORBEL project involving the Advance Light Microscopy, EMBL (Euro-Bioimaging), the Anton Dohrn Marine station, and the Ville-Franche Marine station (both EMBRC), is just one example of the increasing application of machine learning in the life science research. Dr. Rauzi’s team is interested in understanding the mechanics driving tissue morphogenesis during sea urchin embryonic development.

“Using light sheet microscopy to visualize morphogenesis in living embryos is just the first step to study tissue 3D dynamic shape changes” says Matteo Rauzi, Group leader at the Université Côte d’Azur, Nice, France, adding “to extrapolate quantitative information is then necessary to identify objects in 3D (i.e., 3D segmentation) and follow their changes over time. Machine learning algorithms have shown to be powerful tools to perform such a complex task. This course is really a great opportunity for researchers studying morphogenesis to better understand how machine learning works and how it can be applied on complex 4D biological datasets.”

“While the need for such a course was clear from the outset, it was a challenge to decide what the best format for the course would be. We are all excited to present this blended-learning course, which will be a great mix of intensive learning, extensive hands-on and community networking” explains course organiser Tobias Rasse, from the Advanced Light Microscopy Facility at EMBL.

We are very pleased that both NEUBIAS and German BioImaging were interested to co-fund and co-organise the course.

"Machine Learning / Deep Learning is coming to maturity in Life Science and is opening avenues for the analysis of "Big" BioImage Data, but scientists will need to learn how to face the "black-box" magic and how to be critical of it. This course is a pioneer, in that it aims to empower research institutes with the means on how to set up ML/DL and with the knowledge and experience to best use it, and spread it. NEUBIAS intrinsically places knowledge before technology and finds in this course a landmark to follow, for the current integration of Machine Learning in the training of early career scientists in Europe", says Julien Colombelli, Group Leader at the Barcelona Institute of Science and Technology, Spain and Chair of NEUBIAS (COST Action CA15124).



Participants will review the fundamentals of machine learning in three up-front webinars complemented by online tutorials, allowing the participants to get the maximum benefit from the hands-on workshop.

Next, the participants will apply their knowledge at an on-site workshop (EMBL Heidelberg, October 29-31, 2018), in small interactive groups (the workshop has 16 available seats and ~8 trainer/lecturer), to both reference datasets and their own data.

After the on-site workshop, two optional advanced training webinar, complemented by online tutorials, will be given on November 9 and 16, 2018. These will focus on simulation of data, transfer learning and boosting.

After the course participants should be able to:

  • Explain the fundamentals of machine learning methods suitable for image analysis
  • Consult others in strategies to obtain ground truth
  • Give advice in training and using a neural-network
  • Perform simple quality control on the results of one selected ML approach

More information about the course (program & trainers) and venue, are available at the course website.



Registration is open now via the course website. Deadline is June 15, 2018. Applicants should be familiar with Python and should bring their own data. They will be selected based on their skills and letter of motivation. NEUBIAS is providing up to four travel grants for eligible applicants.


About the organisers:

The course is jointly organised by CORBEL, EMBL, German BioImaging and NEUBIAS (COST Action CA15124).

EMBL (European Molecular Biology Laboratory), founded in 1974, EMBL is Europe’s flagship laboratory for the life sciences – an intergovernmental organisation with more than 80 independent research groups covering the spectrum of molecular biology. It operates across six sites: Heidelberg, Barcelona, Hamburg, Grenoble, Rome and EMBL-EBI Hinxton.
…more at

German BioImaging-Gesellschaft für Mikroskopie und Bildanalyse e.V. (Society for Microscopy and Image Analysis, GerBI-GMB) represents the interests of researchers and professionals as well as core facilities in Germany involved in microscopy and image data analysis for the life sciences. The society emerged from the network "German BioImaging" and will also represent the field of biological image data analysis.
…more at

NEUBIAS (COST Action CA15124) is an European network of >220 members and 41 countries, which aims to promote the communication between Life Scientists, Instrumentalists, Developers and BioImage Analysts and to establish and promote the role of Bioimage Analysts in Life Science.
... more at