WORKSHOPS

Speakers:

» Catarina M. Seabra
» Mohamed Edfawy 

Session Description:

In this workshop you will learn how to apply microscopy techniques to answer your questions related to neuroscience research, both in mouse brain and human-derived brain organoids. Using peripheral injections of AAVs expressing a fluorophore under the control of a neuronal promoter, well individualized cells can be found across the entire mouse brain within 4-6 weeks post-injection. We have employed this technique to characterize neuronal arborization, spine morphology and provide proof of principle on its usability towards advanced methods to dissect neuronal network activity or gene function.

The recent breakthrough of the development of brain organoids has now allowed researchers to study the function of brain development and circuit formation in a more physiological context that has a 3D structure. From immunohistochemistry to in-vivo calcium imaging, microscopy techniques have been vital to dissect the intricate processes of brain organoid development and maturation.  

Speakers:

» Arrate Muñoz Barrutia
» Daniel Sage
» Hugo Botelho
» Joaquim Soriano Felipe


Session Description:

» Bridging Deep Learning to ImageJ - Arrate Muñoz Barrutia & Daniel Sage

Machine learning (ML), and in particular, Deep Neural Networks (DNN), have become an inflexion point in many areas of scientific research. In the case of biomedical image analysis, these new techniques provide significant improvements in most of the tasks such as denoising, super-resolution, segmentation, detection, tracking, response prediction or computer aided diagnosis.

Nonetheless, the use of DNN models requires previous programming knowledge and expertise, which makes them unapproachable to the general public. Therefore, the spread of this technology to the scientific community is strongly limited. In this workshop, we present a friendly interface for the use of these models that have been developed in collaboration between EPFL and UC3M. This ready-to-use ImageJ/FIJI plugin has the potential to make available many of the powerful algorithms for image processing that are continuously being developed and published, enhancing research.

 - more info here:

     https://deepimagej.github.io/deepimagej/

     https://www.biorxiv.org/content/10.1101/799270v1

» Analysis of time lapse microscopy data with CellProfiler et al - Hugo Botelho

Automated microscopy is an invaluable tool in the research of complex cellular/organismal processes because it enables reproducible highly multi-dimensional imaging experiments (e.g. multi-position xyλt, where each position may represent a different experimental treatment across a multi-well plate). Scenarios like this are common in live cell assays or high content screening campaigns. Unfortunately, batch image analysis and proper result annotation (experimental metadata) may not be trivial for many researchers. Herein, I propose to demonstrate using CellProfiler in analyzing a small time lapse microscopy dataset with the goal of comparing the phenotypic response of intestinal organoids to gene editing and pharmacological intervention  https://github.com/hmbotelho/cellprofiler-practical2-NeuBIAS). The exercise will focus on how the following basic image analysis tasks are implemented in CellProfiler: background subtraction, object segmentation, feature extraction and results output). A pre-configured analysis pipeline will be provided for the users which wish to follow the demonstration in their own laptops. In the last few minutes I will show how the same
dataset could be analyzed in Knime or Fiji. Participants will not be able to follow along on their own computers (otherwise, the demonstration would be too lengthy) but pre-configured workflows/scripts will be made available as well.

Depending on available time and participant interest, advanced operations (e.g. object-level quality control, tracking and downstream data analysis) may also be demonstrated.

» ImageJ macro-based workflows produce better science


Science is hampered by reproducibility and economic issues. On one hand, lack of reproducibility is due to deliberate data biasing, lack of knowledge, inadequate supervising (usually from IP to Postdocs to students) and incomplete reports. On the other hand, economy limits data production and data sharing to the labs that can pay for: journal access and journal publication expenses, highly specialized tools (i.e. microscopes and image analysis software) and knowledge (i.e. courses and conferences). ImageJ macro-based workflows increase productivity and reproducibility, ease data sharing and are accessible to all users regardless of their technical background and economic status. Practical examples on the topic will be shown.
 

Speakers:

» Pedro Matos Pereira (@P_Matos_Pereira)

» Siân Culley (@SuperResoluSian)


Session Description:

Despite the widespread use of Super-Resolution techniques, most methods remain technically challenging and involve sophisticated hardware and/or software tools. In 2014, Expansion Microscopy (ExM) was introduced as an alternative to overcome this challenge. The concept is extremely simple: instead of super-resolving your target of interest you physically expand your sample in such a way that the structure is now above the diffraction limit (the expansion is 4 to 5 fold the original size) and image it on a conventional microscope. The original method served to deliver super-resolution information to researchers without access to a super-resolution system. However, ExM versatility was limited by technical details like the need for special labelling probes. Several improvements were made since then to enhance its applicability, such as introducing changes that made ExM compatible with conventional probes, like commercially available labelled antibodies and endogenous fluorescent proteins. Additionally, ExM can also be combined with classical super-resolution microscopy techniques effectively allowing to combine the resolution increase from both approaches. ExM has been applied to a variety of samples, from eukaryotic cells and tissues to pathogens such as bacteria or virus, making it an excellent alternative to classical super-resolution approaches.

Speakers:

» Sébastien Tosi
» Ignacio Arganda Carreras

Session Description:

Deep learning, the latest extension of machine learning, has pushed the accuracy of algorithms to unseen limits, especially for perceptual problems such as the ones tackled by computer vision and image analysis. This workshop will cover the foundations of the field, the communities organized around it, and some important tools and resources to get started with these techniques. Some successful applications of deep learning, especially in the field of bioimage analysis, will be presented and some hands-on will cover how to setup a working deep learning environment. No prior programming knowledge is required to follow the workshop.