Short Description

Measures velocities, pressure and forces inside the cell using 2D or 3D video microscopy data



BioFlow is an image-based method able to extract velocities, pressure and forces inside living cells observed in fluorescence microscopy. More details on the method itself are available in the publication linked to this page.

This plug-in provides a ready-to-use implementation of BioFlow for the specific case of a purely viscous model of the cell interior, as demonstrated for the human parasite Entamoeba histolytica (the causative agent of human amoebiasis). However, the code is fully accessible and its great strength lies in its ability to adapt to more theoretical models with minimal effort.

Installation instructions

While we at Icy are proud defenders of the “one-click” install-and-run policy, BioFlow does requires a few third-party libraries to run that are not so easy (or small) to provide “built-in”, therefore we offer here two three alternatives (ranked from simplest to hardest) to get BioFlow up and running on almost any system in no time.

Option #1: Docker (recommended)

As of version 2.0, BioFlow is able to run within a Docker container, which makes everything dramatically lighter and easier (at least for the end-user…). All you need to do is to download and install Docker for your system (here are direct links for Windows and macOS downloads). Start Docker once installed, and that’s it! You can now type “BioFlow” in your favorite search bar and read below for usage instructions.

NB: Docker is used by an exponentially growing number of developers, so this is a good investment of both your disk space and precious time.

Option #2: virtual machine (heavy, but does the job)

In case your machine does not support Docker for whatever reason (age, drivers, bad weather), you can rely on a good’ol’ virtual machine. We have packaged a so-called “virtual appliance” (basically, a linux box containing all necessary software) at the following link (Warning: you’re in for a ~4GB download, compared to a mere 800MB via Docker… you get the picture). VM last updated: Sep 18, 2017.

Similarly to the Docker option, you need one extra piece of so-called “virtualisation” software to install to run this virtual machine. We have succesfully tested it on Oracle’s VirtualBox, although this should equally work on others. Once installed, import the virtual appliance you just downloaded and fire it up. Icy is waiting nicely for you inside with BioFlow installed.

Option #3: manual installation (where simplicity = f(your operating system))

BioFlow relies on a the following thrid-party libraries and software, which should be installed prior to using the Icy module:

  • Python, a popular (though aging*) programming language in the scientific community
  • Fenics is a Python-based Finite Element library at the core of the computations
  • Mshr is a Fenics-based module to generate Finite Element meshes
  • Dolfin is a Fenics-based module used to calculate adjoints.

* Yes, Python dates back to the early 80s…

Installing the software above should be (relatively) straightforward, but depending on the versions (and your OS), your mileage may vary (greatly). This is the list of steps used to pre-configure the virtual machine in option #1 with Fenics version 2016.2.0, based on Lubuntu Zesty Zapus (17.04):

  1. Add the repository for dolfin-adjoint: sudo apt-add-repository ppa:libadjoint/ppa
  2. Update the list of available packages: sudo apt-get update
  3. Install (almost) all required software: sudo apt install dolfin-bin fenics python-matplotlib python-scipy python-dolfin-adjoint
  4. Install Mshr v.2016.2.0 (Fenics’s mesh generator) from source (at cmake/make/sudo make install/sudo ldconfig (NB: Mshr is now bundled as part of Fenics as of version 2017.1)

Running instructions

BioFlow proposes a minimal graphical user interface letting choose to open the 2D or 3D script, which will open in Icy’s script editor. The running instructions are therefore quite simple:

  1. Load up a 2D or 3D sequence into Icy
  2. Open the corresponding script
  3. Define a region of interest around the cell in the first image
  4. Click “run” in the script editor

Extra buttons will let you open up sample data in 2D or 3D to give it a try (basically a pair of consecutive images).

A quick start guide with minimal running instructions and details on the method is also bundled as a PDF within the plug-in, and can be opened via the “Open README” button.

Leave a Review

Leave a review
Cancel review
View full changelog
Close changelog


  • Version • Released on: 2018-01-05 16:21:38

    Added Viewer: BioFlowDisplay

  • Version • Released on: 2017-09-26 21:02:44

    New: BioFlowDisplay module

  • Version • Released on: 2017-09-21 11:33:44

    (Attempt to) fix an intermittent issue causing an "Errno: 20000" when running the BioFlow script

  • Version • Released on: 2017-09-11 18:28:40

    BioFlow nows runs on Docker, hurray!

  • Version • Released on: 2017-08-30 14:57:07

    Compatibility with the virtual machine

  • Version • Released on: 2017-08-29 23:07:37

    Added Dolfin-Adjoint v.2016.2.0 to the list of tried-and-tested versions

  • Version • Released on: 2017-08-29 20:04:05

    Fixed an un-zipping issue

  • Version • Released on: 2017-08-29 19:58:37

    Update to the previous fix

  • Version • Released on: 2017-08-29 19:51:28

    Added index rounding-off to prevent python warnings/errors

  • Version • Released on: 2017-08-29 13:48:12

    Added error reports from the CPython layer

  • Version • Released on: 2017-08-28 19:36:29

    Fixed script access issue (you though v1 would work right away? XD)

  • Version • Released on: 2017-08-28 18:54:06

    First official version of BioFlow, released with the online publication

  • Version • Released on: 2016-11-15 19:12:54