Fluorescence microscopy allows researchers to explore the living world at the cellular and subcellular scales with remarkable precision. However, as time passes, microscopes inevitably degrade: detectors become noisy, optical systems lose alignment, and image quality declines. This aging process can hinder long-term biological studies and quantitative analysis.

To address this challenge, a team of Engineers from IBDM and LIS (France-BioImaging Marseille node) developed μPIX, a new deep-learning algorithm based on generative artificial intelligence.

A smarter way to restore microscopy images

μPIX uses a specific type of AI called a Pix2Pix conditional Generative Adversarial Network (cGAN): this algorithm learns how to transform low-quality or noisy images into clean and high-quality ones, based on examples.

Figure: µPIX architecture is based on a Pix2Pix generative network. µPIX consists of two subnetworks: a generator, based on a UNet architecture with an EfficientNet-b0 backbone, and a discriminator (PatchGAN). During supervised training, a noisy image is input to the generator, which generate an image. This output is compared to the real clean image using a pixel-wise loss function (MSE). Pairs of real and generated images are then passed to the discriminator, which classifies them as real or fake using a binary cross-entropy loss (BCE). Both subnetworks are progressively refined through adversarial loss during training. In the inference phase, only the trained generator is used to generate clean images. (Bon, Gabriel, Sapède, Daniel, Matthews, Cédric and Daian, Fabrice. “μPIX: leveraging generative AI for enhanced, personalized and sustainable microscopy” Methods in Microscopy, 2025. https://doi.org/10.1515/mim-2024-0024)

Unlike conventional image processing algorithms, μPIX adapts its training to the characteristics of the microscope, making it personalized and highly precise.

It improves image quality while preserving fine structures and intensity relationships, which is essential for quantitative imaging.

Thanks to its capacities, it extends the usefulness of old equipment, offering a cost-effective and sustainable alternative to replacement.

Better results than existing tools

In their publication, the authors show that μPIX outperforms both traditional denoising methods and popular deep learning tools such as CARE or Cellpose3

It also improves downstream applications: using μPIX as a pre-processing step enhances segmentation accuracy by up to 3% compared to existing pipelines.

Reviving aging detectors

The team went one step further and applied μPIX to an ambitious task: restoring images from an outdated Multi-Alkali photodetector so they resemble those acquired with a high-performance GaAsP detector.

The results are impressive: μPIX manages to compensate for signal loss along the z-axis (represents the depth), recover structural information, and maintain a near-linear relationship between the predicted and original intensities, enabling quantitative analysis on images that would otherwise be considered obsolete.

From user-centered to hardware-centered AI

Unlike most AI tools that require users to train their own models, μPIX proposes a platform-centered paradigm: platforms train one model, tailored to their equipment, and provide it to their users. This approach reduces redundancy, improves consistency, and aligns with the principles of frugal and shared AI development.

The code and models are freely available on GitLab, and μPIX is already proving to be a useful asset for microscopy platforms seeking long-term performance with limited hardware budgets.

Read the full scientific article here: https://www.degruyterbrill.com/document/doi/10.1515/mim-2024-0024/html

Recently, Pierre Bourdoncle, head of the IMAG’IC platform at the Cochin Institute (Paris Centre Node), and his team published a new protocol for intravital imaging of calvarial bone marrow. Today, he tells us more about their research and how it can enhance the study of diseases like leukemia.

Could you tell us a little about yourself and the project?

As the head of the IMAG’IC platform at the Cochin Institute, we have consistently advanced intravital imaging through multiphoton microscopy. For the past 25 years, we have been dedicated to enhancing intravital imaging at the Cochin Institute, with a focus on improving synchronization, laser technology, and OPO (Optical Parametric Oscillator, ed.) systems.

Why is the calvarial bone marrow such an interesting model to study hematopoiesis and vascular dynamics?

The calvarial bone marrow is an interesting model for studying hematopoiesis and vascular dynamics due to its unique anatomical features. Its thin structure allows for high-resolution imaging, facilitating the observation of cellular interactions and vascular networks. Additionally, it is easily accessible, making it ideal for experimental manipulations and real-time monitoring. This model provides valuable insights into the complex processes of blood cell formation and vascular development.

z-projection of tile scan view of the calvaria vasculature labeled by cdh5-DSRED – 2-photon microscope

Your team has developed a custom-made titanium cranial implant. What advantages does it offer compared to existing methods?

The stability of the imaging area has always been a major challenge in intravital microscopy. Indeed, the animal’s breathing and temperature variations complicate long-term acquisitions. Moreover, precise repositioning of the acquisition area over several days is essential for observing the evolution of the cellular environment. The development of titanium implants, as opposed to traditional resin 3D printing, allows for more robust fixation of the system to the microscope stage and, most importantly, limits the deformation of the implant.

(A) Parts of the implant in situ: 1 observation ring, 2 cementing feature, 3 stabilizing anchor, 4 tail, 5 dovetail, 6 threaded hole, 7 Bregma. (B) Connection of the head implant to the holder: 8 fixation body, 9 clamp, 10 eccentric lever, 11 structure, 12 microscope objective.

What perspectives does this method open for the understanding of hematological diseases, such as leukemia?

This method opens significant perspectives for understanding hematological diseases like leukemia by enabling detailed visualization of disease progression and cellular interactions. It allows researchers to study the impact of treatments in real-time, enhancing the development of targeted therapies. Additionally, it facilitates the exploration of the bone marrow microenvironment’s role in disease pathogenesis.

What are your upcoming projects?

Following the same principle, we are collaborating with the company Ymetry to develop similar appendages adapted for soft organs. Our goal remains to maintain the acquisition area for as long as possible without any drift.

Click here to read the full scientific article!

We’re proud to announce the official integration of our two new cutting-edge imaging Nodes into the French Node of Euro-BioImaging: Normandie and Rhône-Alpes!

With this upgrade, the French Node now spans 10 geographical sites and provides access to 30 state-of-the-art imaging facilities, supporting both national and transnational users.

Why does this matter for the Euro-BioImaging community?

These new Nodes significantly expand the scope and excellence of the French Node:

Normandie Node (PRIMACEN – Rouen) brings unique expertise in:

  • Intravital imaging for vascular diseases, including brain and heart imaging
  • Microalgal biosciences and marine biology imaging
  • Advanced cryo-correlative microscopies (CLEM) and super-resolution imaging

Rhône-Alpes Node (LYMIC – Lyon & ISDV – Grenoble) enhances our capacity with:

  • Pioneering technologies in biomechanics and mechanobiology
  • Rare capacities in spatial transcriptomicsadaptive optics, and metabolic imaging
  • Deep expertise in large-volume 3D EM with integrated image analysis pipelines

These new capabilities fill critical technological and geographic gaps and will benefit users across Europe seeking access to next-generation imaging and expert support.

Users can expect powerful collaborations, robust training opportunities, and access to highly specialized platforms.

The Executive Board of the Rhône-Alpin Node of France-BioImaging is pleased to invite you to the event “Imaging & Microscopy Day in Rhône-Alpes – Image Analysis” – pre-program attached.

It will be held on Tuesday, July 1st at the Faculté Rockefeller, 69008 Lyon

Registration here: https://docs.google.com/forms/d/e/1FAIpQLScTyOV5tjfxZp2i4j-6OmXzdLUcHa7ByMOiJPnBMkEZZkAVBg/viewform?usp=header 


During registration, we will collect proposals for:

  • User Case Presentations (16 + 4 min questions) – Ideal for PIs and established researchers
  • Short Communications (8 + 2 min questions) – Perfect opportunity for students and early-career researchers

If you have any questions, please contact the organizers:

Jacques Brocard: jacques.brocard@ens-lyon.fr
Olivier Destaing: olivier.destaing@univ-grenoble-alpes.fr
Xavier Jaurand: xavier.jaurand@univ-lyon1.fr

As the number of seats is limited, please register as soon as possible to best organize the final program!

We organize in Pasteur a training school on bioimage analysis at the Institut Pasteur, Paris, in May 2025.

The school will be in person only, from the 12th to the 16th of May 2025. All the details are on the course page, some details below.

The course lasts one week and is made of 2 tracks that run in parallel:

  • Early career investigators track (ECI): Learn to master the tools and techniques of bioimage analysis for your own research. From power usage to building analysis pipelines.
  • Analysts track: Learn to use and deploy advanced tools; learn to master high-performance computing for advanced bioimage analysis.

The number of available seats is 25 students max for the ECI track and 15 for the Analysts track. The selection is based on project description.

The keynotes are common to both tracks, and there is a bonus session on Friday afternoon: Work on your own data, with the help of colleagues and experts.

Program

The exact schedule is still being finalized. Here is a description of the course content.

Both tracks of the course have a specific focus on hands-on and interactive tutorials. They are meant to be convivial and foster a collaborative atmosphere between students and teachers. Each day begin with a common keynote, then the program for each track takes place.

Early-career investigator track

In this course you will learn how to use the most recent and common image analysis software tools. You will learn to master and use them for your own research project. The course will walk you from their installation, basic usage to building image analysis pipelines, from raw images to quantification results.

In the beginning we will explore the usage of software such as Fiji, Icy, QuPath, Ilastik, TrackMate, and Deep Learning tools… By the end of the course you will able to use and edit scripts and notebooks for batch processing and some advanced analysis.

The course will also offer fundamental introductions to the topics in modern image analysis, including machine learning / deep learning, ethics, …

You should apply to this course if you are a biologist and / or have no or little background in image analysis and do imaging in your research project. No knowledge of coding is required.

Analyst track

The strong focus of this track is the use of advanced algorithms, and mastering new tools and techniques. For every edition of this course, we pick a central topic in image analysis that we use to articulate the lectures and practical sessions of this track.

This year this topic is image analysis in the scope of spatially-resolved omics. Spatial-omics is a term used to describe a wide range of technologies focused on studying the molecular composition and interactions within tissues or cells while maintaining their spatial context. They all involve imaging and image analysis. We will use spatial omics as a theme to articulate several lectures and practical sessions on advanced image analysis topics that are central to these technologies. Importantly: we will restrict the topics to be on image analysis only, and won’t be dealing with the bioinformatics part. However, guest lectures by experts will help contextualize the course content within the broader scope of spatial omics.

In addition, the course will also focus on the use of artificial intelligence for bioimage analysis, using computational pathology and cell biology as topics to articulate the sessions and lectures.

Finally, a session will be dedicated to high performance computing in bioimage analysis, in the context of large images and large datasets.

The main tools of this track will be Python, Napari and Icy.

Basic experience with scripting and python is required.

Requirements

Bring your own laptop. We will spend time together installing everything needed and making sure they run for the course.

Also, absolutely bring a mouse with the laptop :) It’s painful to use the tools mentioned above with the trackpad.

Participants are encouraged to bring image data for the ‘Work on your own data’ sessions.

Registration

For registration visit the course webpage here : https://www.pasteur.fr/en/education/programs-and-courses/pasteur-courses?id_cours=32420

Deadline for registration: March the 31st 2025

Date for acceptance / rejection communication: April the 3rd 2025

Fresnel Institute, in collaboration with Imaris Software, is organizing the Imaris Workshop Day on Tuesday, March 11th.

This event includes a general presentation on Imaris, during which an Imaris expert will showcase various examples of its applications. Following the presentation, there will be an image analysis clinic where you can discuss the analysis of your own data*.

Workshop program:

  • 13:30-14:30: Imaris presentation
  • 15:00-17:30: Image analysis clinic

Location: Salle Pierre Cotton, Institut Fresnel, Faculté des Sciences – 52 Avenue Escadrille Normandie-Niémen, 13397 Marseille.

Registration is free of charge but mandatory. You can register here or click on the file below.

*If your data isn’t ready by then, we’ll find a similar dataset to discuss.

Gustave Roussy Microscopy Facility, in collaboration with Imaris Software, is organizing the Imaris Workshop Day on Friday, February 14th.

This event includes a general presentation on Imaris, during which an Imaris expert will showcase various examples of its applications. Following the presentation, there will be an image analysis clinic where you can discuss the analysis of your own data*.

Workshop program:

  • 11:00-12:00: Imaris presentation
  • 13:00-16:00: Image analysis clinic

Location: Salle 2, Espace Maurice Tubiana, 20 Rue du Dr Pinel, 94805, Villejuif.

Registration is free of charge but mandatory. You can register here or click on the file below.

*If your data isn’t ready by then, we’ll find a similar dataset to discuss.

L’institut CIML (Centre d’Immunologie de Marseille Luminy), membre du nœud marseillais de France-BioImaging, organise en mars 2025 une formation dédié à la microscopie confocale spectrale.

L’objectif de cette formation, ouverte aux ingénieurs et chercheurs utilisant la microscopie confocale, est d’acquérir en mode spectral un panel 10 couleurs sur coupe de tissu et analyser les images réalisées.

Pour vous inscrire, vous devez préalablement remplir le questionnaire disponible ici: https://france-bioimaging.org/wp-content/uploads/2024/12/Questionnaire_Formation-microscopie-confocale-spectrale-mars-2025.pdf

Retrouvez le programme de la formation ci-dessous:

Vous avez jusqu’au 27 janvier 2025 pour vous inscrire et retourner le questionnaire complété!

Contact:

Hélène Pastor: Chargée de formation et du développement des ressources humaines – Inserm
demat-form.dr-marseille@inserm.fr

Inscription (aucune demande ne sera prise en compte sans le questionnaire):

  • Personnels Inserm ou non Inserm dans une structure mixte Inserm : inscription via www.sirene.inserm.fr + envoi du questionnaire à : demat-form.dr-marseille@inserm.fr (Région : Paca – Domaine : TS3 – Imagerie)
  • Autres personnels : formulaire d’inscription + questionnaire à transmettre à demat-form.dr-marseille@inserm.fr

The NeurImag cellular and molecular imaging Facility, member of the Paris Centre Node of France-BioImaging, has initiated the development of a new tool called ExoJ, in collaboration with the teams of Guillaume Van Niel (CRCI2NA, Nantes University), Frederik Verweij (Utrecht University), Thierry Galli (IPNP, Inserm, Université Paris Cité) and Junjun Liu (Shandong First Medical University).

What is ExoJ?

ExoJ is a plugin developed for the Fiji/ImageJ2 software, specifically designed to automate the reliable detection and analysis of exocytosis events from fluorescence microscopy images. Exocytosis is a cellular process where molecules or substances contained within a cell are released to the extracellular environment. This process involves the fusion of a vesicle, a membrane-bound sac, with the cell membrane. Once fused, the contents of the vesicle are expelled into the extracellular space.

How does ExoJ work?

ExoJ automatically identifies user-defined exocytosis events. It extracts key quantitative information such as the intensity, apparent size and duration of each event. ExoJ is fully parameterizable and configurable, making it suitable for studying different types of exocytosis, whatever the imaging modality (TIRF [1] and/or spinning disk [2]). ExoJ is a robust and reliable tool for analyzing large datasets!

What are the benefits of ExoJ?

ExoJ automates the detection of exocytosis events, considerably reducing analysis time compared with manual annotation. Moreover, the results obtained are reproducible, facilitating comparisons between different experiments. Finally, ExoJ is based on Fiji/ImageJ2, an open-source software widely used in the scientific community.

To read the article, click here.

[1] Cois et al., 2024 https://pubmed.ncbi.nlm.nih.gov/39145986/

[2] Hessvik et al., 2023 https://pubmed.ncbi.nlm.nih.gov/37285022/

From February 6 to 7, 2025, the University of Rouen Normandie will host the 8th edition of the France Cerebellum Club Days. This year’s event will include a session dedicated to cerebellum bioimaging, highlighted by the Primacen imaging platform, a member of the Normandie Node of France-BioImaging.

The France Cerebellum Club is an organization aimed at promoting exchanges between scientists involved in the study of the cerebellum in all its modalities, using a variety of analysis methods.

This new edition will bring together researchers and industrials to discuss the latest advances in the study of the cerebellum. On the program: plenary lectures, thematic sessions and workshops highlighting recent work on the development, functions and pathologies associated with this cerebral region. This year’s topics include the development and evolution of the cerebellum, innovations in applied bioimaging, organoid models and studies of connections between the cerebellum and other brain regions.

Two keynotes will surround these scientific days. Mari Sepp (Heidelberg, Germany) will present her work on the development and evolution of the cerebellum using single-cell genomics, while Christian Hansel (Chicago, USA) will discuss cerebellar instructive signals and their role in neocortical plasticity.

For more information and registration details, click here.

Le 3 décembre dernier, le Nœud Paris-Centre, membre de l’infrastructure France-BioImaging a organisé une journée intitulée placée sous le signe de la valorisation scientifique et du transfert de technologie. Accessible en mode hybride, elle a réuni une dizaine d’intervenants et une quarantaine d’auditeurs.

L’objectif de cette journée était de rendre accessible la valorisation de l’activité de recherche au plus grand nombre de chercheurs, étudiants, et acteurs du transfert technologique. À cette fin, les participants ont pu bénéficié d’une série de témoignages de chercheurs, de directeurs de startups (ABBELIGHT AVATAR MEDICAL, TWINCKLE FACTORY, INSCOPER) de responsables d’entreprises, et d’acteurs des services de valorisation qui ont converti un résultat scientifique dans le domaine de l’imagerie biologique en un produit ou un service commercialisé.

Ces témoignages ont permis aux intervenants de faire part de leur motivation et leur parcours de transfert technologique avec ses difficultés et ses satisfactions. Les directeurs de startups et responsables d’entreprises ont également présenté aux participants les temps clés de la valorisation d’une découverte scientifique, tels que les échanges entre industriels-chercheurs et la mise en place de la valorisation économique du travail scientifique initial.

Pour clore cette journée, une la table ronde finale a permis aux acteurs des services de valorisation de décrire la fonction, les moyens, ainsi que les modalités d’interaction avec les chercheurs et d’accompagnement du transfert technologique.

Restez connectés, les certaines sessions seront bientôt disponibles en replay!

On September 6, the Cellular Imaging Master, created in 2004 by Delphine Burel and Ludovic Galas at the University of Rouen Normandie, celebrated its 20th anniversary in partnership with the IBiSA PRIMACEN imaging facility.

Over 50 alumni attended the event, along with the Master’s coordinators and several teaching staff members. The day began with a review of the program’s 20 years and a presentation on international collaborations with the Universities of Turku and Abo Academy in Finland, followed by a vote for the new Master’s logo!

In the afternoon, participants enjoyed fun activities such as karting, bowling, and karaoke at Espace Loisirs Rouen, fostering a convivial atmosphere among students and teachers. Everyone agreed to meet again in five years for future celebrations.

This successful anniversary was made possible thanks to the efforts of the M2 IMAC 2024 class and the support of their sponsors!

Find out more about the IMAC Master’s program here.