BioImage Data Analysis Course

(BIAS 2)

The 1st edition of the BIAS course was organized by Kota Miura, Sébastien Tosi and Christoph Moehl in May 2013 at EMBL (Heidelberg, Germany).


This second edition is intended for Image analysts, i.e. specialists working in European Imaging Facilities. Participants attendance is limited to 25, therefore participants will be selected by the organizers.

The participation is subject to a 500 euros fee which covers participation to the Community Meeting (2 previous days), lunch, coffee breaks, a limited computer rental and a contribution to the traveling of teachers.

To register, please fill in the application form.

Teachers / Instructors

    Kota Miura, Centre for Molecular and Cellular Imaging (CMCI), EMBL Heidelberg.    - Organizer -

    Sébastien Tosi, Advanced Digital Microscopy, IRB Barcelona.                                - Organizer -

    Cristoph Moehl, DZNE, Bonn.                                                                              - Organizer -    
    Julien Colombelli, Advanced Digital Microscopy, IRB Barcelona.   

    Thomas Pengo, Center for Genomic Regulation, Barcelona.                                   

    Perrine Paul-Gilloteaux, Institut Curie, Paris.                                                        

    Christian Tischer, Advanced Light Microscopy Facility, EMBL Heidelberg.        

    Simon Nørelykke, ETH Zuerich.

    Ricard Delgado Gonzalo, EPFL Lausanne.                                                                    


The Course will be organized after the Community Meeting, from Wednesday 9th until Saturday 12th (included) of October 2013 (see course planning for details). Participants are strongly recommended to attend to the community meeting on October 7-8th.


    Introduction to ImageJ macro programming

Writing ImageJ macros for efficient automatic image processing and analysis.
  • Macro editor and recorder.
  • User interactions: getNumber, getString and Dialog boxes
  • User interruption: waitForUser
  • Variables
  • User-defined functions, passing arguments
  • Calling a plugin from a macro
  • Loops and conditional execution
  • File browsing and batch processing
  • Some simple example macros
    Overview of ImageJ scripting and programming environments

Introduction on how to use plugins that do not support macro scripting, and how to build your own plugins.
  • The limitations of the macros of ImageJ
  • Scripting as alternative to macros
  • Build your own plugin
  • Pixel-wise operations
  • What is the ImageJ API, and how to use it
  • Using plugins that are not compatible with macros 


    Introduction to programming and data analysis with Matlab

Introduction to the Matlab programming environment and basic programming concepts. Include small applied projects to get familiar with Matlab image processing functions.
  • Matlab environment: command line, workspace, script editor
  • Variable types
  • Working with matrices
  • Basic calculations
  • Loops
  • Importing images and visualizing images
  • Image normalization
  • Plotting the grey value histograms of raw and normalized images
  • Detecting image objects by thresholding and connected components analysis
  • Calculation of image object features
  • Visualization of different image object features by scatter plot in 2D parameter space

Application projects

            FISH spots counting in human spermatozoids


FISH (Fluorescent In-Situ Hybridization) is a complex gene staining technique with numerous variants. The quality of the staining depends on related physical parameters such as the level of DNA de-condensation. We will essentially consider here that our protocol allows to label some chromosomes of interest inside human spermatozoids so that
they appear as "bright fluorescent spots". The students will write an ImageJ macro processing images from such FISH assay to segment the spermatozoid nuclei and classify them based on their chromosomal content (multi-
plicity of FISH spots in 3 different fluorescent channels). The workflow will also include some results visualization for simple assessment of the classification quality and statistics computation.

    Input data

The nuclei of the spermatozoids are DAPI stained and the chromosomes of interest (here X, Y and 18) are FISH stained. The fixed sample is scanned by a motorized stage microscope (widefield or confocal), all the fields of
view necessary to tile the sample are acquired.

            Statistical Analysis of Microtubule Orientation


Regulation of cytoskeletal orientation is a basic mechanism for controlling cell polarity. Using ImageJ and Matlab, we explore strategies for quantifying EB1 movement along microtubule within cultured cells. Students learn how to measure directionality of movement in image sequences by a combined use of tracking tools in ImageJ and statistical analysis using Matlab.

    Input Data

2D time series of a single cultured cell, EB1 labeled.

            Quantitative evaluation of muticellullar movement in Drosophila


In this module the students will be instructed how to track cell movements within the mono-layer sheet of the epithelium of a Drosophila embryo. The segmentation of the cells is performed by an ImageJ macro on the maxi-
mum intensity projection of the original movie, the macro outputs a stack of segmentation masks (one mask per time frame). The tracking is performed in Matlab by importing the segmentation masks. The students will also learn how to identify the junctions (cell-cell adhesions) and the vertices (intersections of junctions), and to visualize the results of the tracking.

    Input data

The cell membrane expresses a GFP-E-cadherin construct, the tissue of the embryo is imaged in its apical part by a spinning disk microscope. The datasets are provided as the maximum intensity projection of three time-lapses show-
ing different tissue morphologies and dynamics.

            Cell migration polarity: automated classification of dynamics of actin cortex components


Cell migration is driven by polarized dynamics of the actin cytoskeleton and connected focal adhesion sites. In the following project, we learn how to classify detected image objects (focal adhesions) due to different object features (amongst others: actin flow above focal adhesions) by data mining techniques.

    Input data

  1. 6 min time-lapse series of a migrating cell in low temporal resolution. Focal adhesion are imaged in epifluorescence
  2. 70s time-lapse series of the same migrating cell in high temporal resolution, imaged immediately after movie 1. Actin is imaged in TIRF.

            3D visualization and quantification of Blood Vessels inside a Subcutaneous Tumour


For this project mice developing some specific tumors are injected a rhodamine-lectin construct to stain their blood vessels before sacrificing. Extracting the local statistics of the blood vessel network inside specific sub-regions of the tumors is a powerful investigation tool: the density of the vascularization and vessel branching points and the thickness of the vessels are all crucial age indicators to understand how the structure developed and possibly necrosed. With the help of a simple ImageJ macro these statistics can be extracted and the network 3D rendered with judicious color/transparency to provide insights on its organization.

    Input data

The datasets were acquired by a custom macroSPIM (Single Plane Illumination Microscope) allowing to image large (up to 1 cm), fixed and optically cleared samples (pieces of organs, tumors, whole organisms...). 3D stacks of different sub-regions of a subcutaneous tumor (lectin stained blood vessels) are acquired.