A new version of TrackMate is available now, with major changes that improve its versatility. TrackMate now integrates state-of-the-art segmentation algorithms from machine-learning and deep-learning such as StarDist, Ilastik and Weka.

TrackMate[1] is a Fiji plugin that address cell or organelle tracking in Life-Science microscopy images. Its main goals are to be user-friendly, interoperable and to serve as a platform to accelerate the development of novel tracking algorithms and analysis pipelines.

With this new version we rewrote almost entirely TrackMate so that it can integrate state-of-the-art segmentation algorithms and benefit from their output. For instance, TrackMate can now store, display, save, load and exploit object contours in 2D.

We also made a new application programming interface that will facilitate and accelerate reusing TrackMate in other analysis pipelines and allow 3rd party contributors to add new segmentation algorithms in TrackMate in an easy way. We used this API ourselves to add 7 new segmentation algorithms to TrackMate:

For instance, the StarDist[2] algorithm is integrated as two different detectors. The first one uses the built-in deep-learning model that can segment cell nuclei in fluorescence image in a wide range of situation. The robustness of the StarDist algorithm in turn positively impacts the robustness of tracking and allows for better detection of cell divisions with TrackMate tracking algorithms. This will facilitate cell migration studies.

The TrackMate StarDist integration also allows for specifying and using a custom deep-learning model. For instance, we trained a specific model to detect T-cells imaged in bright-field microscopy and track them over time. Before the emergence of such detection algorithms, the tracking of label-free cells was difficult.

We also integrated the ilastik[3] segmentation software. A TrackMate user can input an ilastik classifier to detect objects then track them. We used them to study the bacterial growth of Neisseria meningitidis clones. The output of this analysis pipeline offers the lineage of each single cell along with its morphology and how it evolves across cell divisions.

The new capabilities of TrackMate can be used to address applications beyond tracking. For instance, it is now possible to use TrackMate to perform the segmentation of 3D objects using a slice-by-slice approach. This approach consists in segmenting objects in each 2D section of a 3D stack, then merging the segmentation results along Z in a subsequent step. This can be done in TrackMate, using the tracking step for merging. We implemented a novel tracking algorithm to foster this application, the overlap tracker. We could use this approach combining the cellpose[4] algorithm in 2D to segment 3D images of Arabidopsis thaliana floral meristem.

There are several other algorithms that are now offered to the TrackMate user, within a user-friendly software meant to interoperate with the key software of bioimage analysis. More importantly, TrackMate is an open-source academic software, and its new API will foster the development of new analysis pipeline with TrackMate and the integration of new algorithms by other developers, increasing the breadth of applications it can address for Life-Science researchers.

This new version of TrackMate is the product of a collaboration between the IAH facility (Institut Pasteur), part of the FBI Bioimage Informatics Node , the Jacquemet lab (Turku Bioscience Centre) , and the Dumenil lab (Institut Pasteur) .

Bringing TrackMate in the era of machine-learning and deep-learningDmitry Ershov, Minh-Son Phan, Joanna W. Pylvänäinen, Stéphane U. Rigaud, Laure Le Blanc, Arthur Charles-Orszag, James R. W. Conway, Romain F. Laine, Nathan H. Roy, Daria Bonazzi, Guillaume Duménil, Guillaume Jacquemet, Jean-Yves Tinevez bioRxiv 2021.09.03.458852; doi: https://doi.org/10.1101/2021.09.03.458852

Contact: Jean-Yves Tinevez

[1] https://imagej.net/plugins/trackmate/

[2] Alejandro F. Frangi, Julia A. Schnabel, Christos Davatzikos, Carlos Alberola-López, and Gabor Fichtinger Uwe Schmidt, Martin Weigert, Coleman Broaddus, and Gene Myers. Cell detection with star-convex polygons. In Alejandro F. Frangi, Julia A. Schnabel, Christos Davatzikos, Carlos Alberola-López, and Gabor Fichtinger, editors, Medical Image Computing and Computer Assisted Intervention – MICCAI 2018, pages 265–273, Cham, 2018. Springer International Publishing. doi:10.1007/978-3-030-00934-2_30.

[3] Stuart Berg, Dominik Kutra, Thorben Kroeger, Christoph N Straehle, Bernhard X Kausler, Carsten Haubold, Martin Schiegg, Janez Ales, Thorsten Beier, Markus Rudy, Kemal Eren, Jaime I Cervantes, Buote Xu, Fynn Beuttenmueller, Adrian Wolny, Chong Zhang, Ullrich Koethe, Fred A Hamprecht, and Anna Kreshuk. ilastik: interactive machine learning for (bio)image analysis. Nature Methods, 16(12):1226–1232, 2019. ISSN 1548-7105. doi:10.1038/s41592-019-0582-9.

[4] Carsen Stringer, Tim Wang, Michalis Michaelos, and Marius Pachitariu. Cellpose: a generalist algorithm for cellular segmentation. Nature Methods, 18(1):100–106, jan 2021. doi:10.1038/s41592-020-01018-x.

France BioImaging primary mission is to develop, promote, disseminate and provide access to innovative instruments and imaging technologies in the field of bioimaging to scientists. Fostering the technological transfers is at the heart of this mission, and for this France BioImaging relies on a strong association of leading R&D research teams with core facilities.

However, several bottlenecks exist and often hamper or prevent successful technology transfer:

  • A lack of human resource leads to difficulties in transferring and stabilizing the technology which is not enough user-friendly
  • A technology that is too specific, with not enough user base
  • A difficulty to contract with industry through institutional offices for industrial valuation
  • In the context of image analysis: the instability of open software economical model, inter-operability, large data handling and algorithm complexity

As a way to tackle these bottlenecks, France BioImaging launched in January 2021 its first “FBI Internal Call 2021: Technology transfer from the R&D teams to the core facilities” to promote the transfer of new technologies (instrumentation, probes, staining methods, software, data analysis or data visualization) from the R&D teams to the facilities of France BioImaging, for access and service to end-users. The outcome of the transfer project had to ensure for the prototype to be usable by the end-users until the interpretation of the data. The project had also to include a sustainability plan and a training plan to guide both facility staff and end-users toward autonomy.

The project selection was organized by the National Coordination of France-BioImaging and applications were assessed according to the following evaluation criteria:

  • Innovation and originality of the proposal
  • Scientific quality, implementation, timeline
  • Competitive positioning
  • Adequacy of resources with the proposed project
  • Economic impact and tech transfer potential and perspectives
  • Estimation of the user market and potential for user adoption
  • Plan for training and sustainability.

For the first edition of the “FBI Internal Call 2021: Technology transfer from the R&D teams to the core facilities”, 5 projects were selected:

  • Icy@FBI: Jean-Christophe Olivo-Marin (IPDM Node): Broadening the scope of applications of Icy (http://icy.bioimageanalysis.org/) by implementing several key new bioimage analysis components
  • BIC-HCS-SMLM: Jean-Baptiste Sibarita (Bordeaux Node), Technological transfer of a Single-Molecule-based High Content Screening platform to the Bordeaux Imaging Center
  • CloudFISH: Marcello Nollmann (Montpellier Node), A tool for the analysis of single-molecule RNA and DNA FISH images
  • MorphoNet: Emmanuel Faure (Montpellier Node), An interactive online morphological browser to explore complex multi-scale data
  • BioImageIT (https://bioimageit.github.io/#/about): Jean Salamero, Sylvain Prigent (IPDM Node), An open source framework for integration of image data management with analysis

Each selected project was awarded with a 80k€ grant for salary and/or equipment, and several positions are currently available: https://france-bioimaging.org/jobs/

This call will be renewed in 2023.

NEUBIAS is a COST Action which brings together life-scientists, microscopists, bioimage analysts and image analysis developers from 36 European, three neighboring countries + Australia, Singapore and the USA (www.neubias.org).

NEUBIAS is a forum to exchange the newest findings, applications, and cutting-edge developments in Bioimage Analysis, machine learning, data mining, and storage. European Bioimage Analysts, an emergent group within the bioimaging analysis community, organize this event, bringing together an international, interdisciplinary community of scientists in life and computer sciences.

Andreas Girod and Aymeric Fouquier d’Hérouël will be hosting the conference in Luxembourg, which will include a Training School for Early Career Investigators, a Training School for Bioimage Analysts, a Taggathon to continue building the NEUBIAS online resources for the Bioimage Analysis Community. Moreover, the Bioimage Analysis Symposium will be organized from the 6th to the 8th of February, 2018, which will include a new Satellite workshop open for bioimage analysts on the 5th of February (afternoon).

The symposium will highlight Keynote lectures from Susan Cox, Kevin Eliceiri, and Ivo Sbalzarini and will include talks from other 14 exciting invited speakers. Also, contributed talks will be selected from abstracts. The NEUBIAS symposium will feature signature sessions: the Call for Help or “image clinics” session (C4H), the Open source Software Lounge (OsSL), the Panel Discussions as well as company Workshops and Digital Posters.

CORBEL, EMBL, German BioImaging and NEUBIAS are delighted to announce a joint blended learning course on Machine Learning for Image Analysis.

The course will be a great mix of intensive learning, extensive hands-on and community networking. Participants will review the fundamentals of machine learning in three up-front webinars complemented by online tutorials.

The webinars will take place on 2nd, 9th and 16th October 2018, 12:00 – 14:00 CEST but a recorded alternative can be provided.
Next, they will apply their knowledge on-site (EMBL Heidelberg, 29-31st October), 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 9th and 16th November 2018. These will focus on simulation of data, transfer learning and boosting.

NEUBIAS (COST Action CA15124) is providing up to four travel grants for eligible applicants.

Application deadline: June 15th, 2018.

Perrine Paul-Gilloteaux, bio-image analyst, CNRS research engineer and project manager of our Bio-image informatics node, received last month the 2017 CNRS Crystal Prize, awarding her contributions to French research.

A perfect occasion to highlight her career and her work with France BioImaging. What is eC-CLEM? How can our field deal with the massive amount of data produced? What future developments can we expect in the realm of bio-image informatics? Read the interview below to find out more.

Perrine Paul-Gilloteaux
Perrine Paul-Gilloteaux

Could you introduce yourself briefly?

My name is Perrine Paul-Gilloteaux, I’m a CNRS Research Engineer. I have a background in electrical engineering, signal and image processing, and did my PhD in augmented reality for neurosurgery through surgical microscope. I started working in bio-image analysis for microscopy in Ireland, and joined the Curie Institute on the PICT IBISA facility in 2010. I moved to Nantes in 2015, and now work in a biomedical research institute. I am also the project manager of the France BioImaging node Bio-Image Informatics IPDM (Image Processing & Data Management), and work closely with the national coordination on the aspect of data management.

I define myself as a bio-image analyst, meaning that I do my best to bridge the gap between microscopy, image analysis and biology. This means that I’m involved in data management, data processing and data analysis projects, that I provide as a service in facilities or work on as research topics.

How long have you been involved with FBI, and what main projects have you carried out with us?

I’ve been involved in FBI from its inception. I started working within the transversal IPDM working group, where we first defined the state of our management systems and worked on the interoperability of our data bases. I managed the setup of the Curie Image Database, supported by France Bio Imaging, based on the OpenImadis system. In 2015, I was nominated project manager of the IPDM node, led by Jean-Christophe Olivo-Marin and Charles Kervrann. One important part of my mission is to work with the national coordination on the data management aspect in FBI. For this, we started by making a survey of resources and management system on site. This question of data management is now central, and FBI collaborates with other infrastructures at the European level: EuroBioImaging and ELIXIR, but also at the national level with other national infrastructures in biology using microscopies, and with the French Institute of BioInformatics (ELIXIR French node).

You have developed a software called eC-CLEM. Could you explain what it consists of?

For this project, I’ve worked closely with Xavier Heiligenstein (Curie Institute, FBI working group on multimodal imaging). Ec-CLEM (for Easy-Cell-Correlative Light to Electron Microscopy) is a software designed to help correlative microscopies. The purpose is to help the fusion of information obtained by different modalities of microscopy on the same sample (for example electronic microscopy, photonic microscopy, atomic force microscopy, etc.). The software allows to register, i.e. align in the same system of coordinates, multidimensional images with big scale and resolution differences, either with a manual input of the user, either automatically when possible. In addition, it provides an estimation of the error of alignment, based on statistical methods, and detects the deformations that the sample may have undergone. I’ve developed a set of algorithms implemented as plugins for the ICY platform. [note: Perrine has published a paper about eC-CLEM in Nature Methods] During the development of this set of tools, I was greatly helped by the ICY coding parties (Hackatons) organized at Pasteur with the support of FBI, and I would encourage developers to attend such events, as there are always new things to be learnt.

Perrine Paul-Gilloteaux presents her work during the Crystal Prize Ceremony.
Perrine presents her work during the Crystal Prize Ceremony
at Nantes University in December 2017.

Bio-image informatics have taken the center stage lately, as more and more people realize how crucial image processing is for research. Could you expand a bit on that?

It is entirely true, and this is the reason why FBI has had a transversal node on that activity since its creation. I’ve cofounded a network of bio-image analysts (NEUBIAS) for that exact purpose also. The size and the number of data to be processed, the large amount of different questions to be answered from imaging and the interplaying between acquisition and processing to generate imaging data and analysis data, have led to a complexity of analysis which requires expert tools but also expert people. Bio-image informatics is a field of research by itself, and it is now recognized as such. It is bridging the gap between image processing research and biology research based on imaging. It can be seen also from the recent Nobel prizes in chemistry: in cryo-tomography or in super resolution light microscopy, both developments were relying on image processing as an essential part.

What are going to be, according to you, the next big steps and developments in the realm of image processing and data management?

The novelties in our field is two-sided: from one side we have data exploding in size and number, and on the other side, machine learning -and in particular deep learning- benefits from progress in hardware and opens the way to big progress in analysis and in particular in feature recognition (segmentation and tracking).

Regarding data management, the big issues to be solved need to involve the whole imaging community, but also to seek expertise from other fields with the same problems. Technical solutions, both from the software side (with management software such as OpenImadis, Omero, Bioemergences in France BioImaging nodes) or the hardware side (optimized hardware systems, optimized protocols of data transfer) are on their way, but will not be useful if the biologists do not put effort in data curation and data selection.

Up to this day, even with machine learning, tuning a software or a protocol to respond to a particular problem and a particular set of data requires a lot of effort, either to set up the learning network or train it in the case of machine learning, either to combine algorithm for a specific question in adapted workflows or to develop more performing algorithm. It means that we need well trained expert able to master both the image processing aspect and the biological questions behind.

On what will your FBI working group focus in 2018? What can we expect from you? (in terms of new developments, priorities, events etc.)

The priority is definitely to deal with the explosion of data we are facing. In addition to the directions exposed in the previous question (software and hardware solutions), one direction taken by IPDM on data management is the definition of quality data. For this multi-faceted topic, we have already started to set up tools to measure the quality of the data produced in term of resolution for example, based on the expertise in metrology of our facilities members, that we want to demonstrate in 2018. We will also concretize the collaborations between FBI and the other national infrastructures by running tests, for example on the speed of data transfers between node, in order to make sure that at the end of 2018, each user of the FBI nodes can easily access and process her/his data from anywhere.  A technical catalogue of software and hardware resources is under construction, to allow FBI nodes and beyond to benefit wisely from the tools and networks created by FBI. In the first semester of 2018, we will be organizing an event to discuss and define the changes in bio-image informatics that deep learning could bring about (further information to come soon, please refer to the FBI site).

References:

http://www.cnrs.fr/fr/recherche/prix/cristal.htm (in French)

Relevant publications:

Paul-Gilloteaux, Perrine, Heiligenstein Xavier et al. “EC-CLEM: flexible multidimensional registration software for correlative microscopies.” Nature Methods, vol. 14, no. 2, 2017, pp. 102–103., doi:10.1038/nmeth.4170. https://www.nature.com/articles/nmeth.4170

NEUBIAS, the network for bioimage analysts, is organizing a second training school for facility staff. Training is open for all staff scientists, graduate students, post-docs or faculty who work in the context of Bioimaging facilities and provide assistance and training to users in need of Bioimage Data Analysis.

Topics, requirements, program and more: http://eubias.org/NEUBIAS/training-schools/staff/ts5-gothenburg2017/

Extended deadline for abstracts submission: May 26, 2017

The course will be an introduction to image processing and analysis, with a focus on biologically relevant examples. The attendees will learn the fundamentals of image analysis including how to do basic Macro programming in Fiji (ImageJ) for automated batch analysis of images, use different software solutions for image analysis, and will be introduced to visualisation and explorative data analysis after extraction of numerical data from images.

The teachers are Simon Nørrelykke and Szymon Stoma (IDA-ScopeM, ETH, Zurich, Switzerland), and Chong Zhang (SIMBioSys, University Pompeu Fabra, Barcelona, Spain).

The course is suitable not only for beginners in image analysis with no experience, but also for those who want to extend their knowledge of basic principles and more specialised tools.