Analyse d'images

Déconvolution

   La déconvolution est une opération mathématique utilisée pour corriger les images après acquisition.

   En effet, l'image d'un objet est déformée par les distorsions optiques (diffraction, aberrations).

   Cette distorsion spécifique du système optique utilisé, est appelée "Point Spread Function" (PSF).  

   L''image obtenue est le résultat d'une convolution entre l'objet et la PSF du système optique.

   Le processus de déconvolution permet donc à partir de d'une image acquise en 3D, d'utiliser une PSF

   expérimentale ou une PSF théorique, afin de restaurer une image de meilleure résolution. 

   Le logiciel HUYGENS est disponible sur la plateforme MICA pour déconvoluer vos images à l'

et à l'

Analyse et représentation 3/4D

L'analyse d'images 3D est l'étude de volumes reconstruits à partir de piles d'images de coupes optiques.  

 Ces logiciels vous permettront à la fois de visualiser des reconstructions 3D de vos échantillons et de les analyser:

• Segmentation 3D
• Comptage d'objets
• Co-localisation
• Analyses morphologiques
• Suivi d'objets en 3D

Morpheme softwares

Macros programs ImageJ/FiJi

General use of macros in ImageJ / FiJi : download the file .txt or .ijm to your local plugins folder (Fiji.app\Plugins or ImageJ\Plugins) and restart ImageJ. The corresponding macro will appear in the Plugin menu. To use it : you just need to select it in the menu To modify it : open the file with ImageJ.

• Wound Healing assay surface measurements : 2 Macros
  

Keywords : biology : Wound healing, confluence, cell migration / image analysis : surface measurement, kinetics, 2D

Input : Phase contrast images

Goal : Measure the migration of the cells through a surface decrease during the closing of the cellular mat

   1) Macro to be dowloaded : Surface_blessure

Author : F. Brau / IPMC

Date : June 2013

Input format : .TIFF or .STK image stack openned on ImageJ

Test images : Stack wound healing

What does Surface_blessure.ijm do?

In Images

In a few words

A light intensity correction is applied to the image stack, followed by an edge detection (Find edge), then a low pass filtering  to blur the cellular mat. Segmentation of the empty space is done by manual thresholding based on stack histogram. This surface is then measured with the exclusion of small objects : Min area measured at t+1 = 0.01*Min area measured at t. A S=f(t) plot and bi-exponential fit (speed can be calculated ) is given to the user with an image displaying the analsyis done (Blue : surface thresholded / Red : surface measured / Green : original)

_________________________________________________________________________________________________________________________________

   2) Macro to be dowloaded : Wound_healingv05

Author : M. Gesson / C3M

Date : 2018

Input format : Images from Incucyte system stored in a folder

What does Wound_healingv05 do?

In a few words

An unsharp mask followed by a minimum filtering is applied to the image stack. Segmentation is done by a fixed threshold and a filtering on a size criterion to give the surfaces through measurements of ROIs obtained. This macro analyses automatically all the images stored in a folder.

_________________________________________________________________________________________________________________________________

Morpheme softwares

• Obj.MPP (Object detection using a Marked Point Process (in Python 3)): Documentation available on a dedicated page.
  Available at: http://gitlab.inria.fr/edebreuv/Obj.MPP
• SMLM-CEL0 (Single Molecule Localization in Microscopy – Continuous Exact L0): Single molecule localization in microscopy based on a deconvolution algorithm with a L0-regularization term to promote sparsity. The continuous exact L0 (CEL0) functional is minimized using an iteratively reweighted L1 method (IRL1). This software has been tested within the SMLMS 2016 software benchmarking.
  Available at: http://github.com/esoubies/SMLM-CEL0
• SPADE (Small Particle Detector): Python software for detecting a collection of small particles in an image. It is based on a marked point process modeling where the objects belong to a predefined dictionary of shapes. Originally, it has been developed within the ANR project RNAGRIMP for detecting granules in cell cytoplasms.
  Available at: http://pypi.python.org/pypi/small-particle-detection and http://gitlab.inria.fr/ncedilni/spade.
• timagetk (Tissue Image Toolkit): Python package dedicated to image processing of multicellular architectures such as plants or animals, and intended for biologists, modelers and computer scientists. Morpheme has contributed to its development by providing a number of image processing tools.
  Available at: http://timagetk.readthedocs.io