Short Description

Automatic registration in 2D or 3D based on detection or binary mask.

eC-CLEM: flexible multidimensional registration software for correlative microscopies P. Paul-Gilloteaux, X. Heiligenstein et al. Nature Methods 14, 102–103 (2017) : http://www.nature.com/nmeth/journal/v14/n2/full/nmeth.4170.html

Documentation

Source code available on GitHub.

Please note that you should use java 8 to use this plugin (simply upgrade you java installation, nothing to be done from ICY itself). Check the Icy console message after installation to see your java version.

Also look to EC-CLEM plugin

An example of Visual Protocol using this plugin is given here

Tutorials:

Datasets and tutorial are available for some example (real data or synthetic data). Each table below presents the input data: source (likely elecron microscopy data) and target (likely light microscopy) in first raw. Second raw is the source image positionned in target, followed by target image positionned on source. Please do not reuse without permission.

Source Target
Input data Source Simulated Spot LM PDF Step by Step introduction and associated dataset (do not reuse without permission)
Data registered using AutoFinder Source Simulated Spot LM Detection Parameters:

Source spot detection scale 2 (about 3 pixels); Target: manual selection;

AutoFinder Parameters:

Find small part in bigger field of view Reverse
; 1 micron; 90%

 

Source Target
Input data Source nuclei PDF Step by Step introduction and associated dataset (do not reuse without permission)
Data registered using AutoFinder Source Simulated Spot LM Detection Parameters:

Source: spot detection scale 4 (about 13 pixels); Target: detection scale 3 (about 7 pixels) ;

AutoFinder Parameters:

Find small part in bigger field of view
; 10 microns; 70%

 

Source Target
Input data Source Simulated Spot LM PDF Step by Step  and synthetic associated dataset (do not reuse without permission)
Data registered using AutoFinder Source Simulated Spot LM Detection Parameters:

Source and target: ConvertBinarytoRoi 0.2 microns

AutoFinder Parameters:

Find small part in bigger field of view
; 1 micron; 70%
Source Target
Input data Source Simulated Spot LM PDF Step by Step  and synthetic associated dataset (do not reuse without permission)
Data registered using AutoFinder Source Simulated Spot LM Detection Parameters:

Source and target: ConvertBinarytoRoi 3 microns

AutoFinder Parameters:

About the same content, 10 microns 50%
Source Target
Input data Source Simulated Spot LM PDF Step by Step  and synthetic associated dataset (do not reuse without permission)
Data registered using AutoFinder Source Simulated Spot LM Detection Parameters:

Source and Target: spot detection scale 2 (about 3 pixels);

AutoFinder Parameters:

Find small part in bigger field of view
; 1 micron; 70%
Source Target
Input data Source Simulated Spot LM PDF Step by Step  and synthetic associated dataset (do not reuse without permission)
Data registered using AutoFinder Source Simulated Spot LM Parameters: convertbinarytoroipoints 0.2 microns

Autofinder: find small part in bigger field of view

: 1 microns, 90%

 

Source Target
Input data Source Simulated Spot LM PDF Step by Step  (also demonstrating  the use of protocols) and associated dataset (do not reuse without permission)
Data registered using AutoFinder Source Simulated Spot LM Detection Parameters:

Target spot detection scale 3 (about 7 pixels); Source: spot detection scale 5 (about 25 pixels);

AutoFinder Parameters:

Find small part in bigger field of view
; 1 micron; 70%

 
Source Target
Input data Source Simulated Spot LM PDF Step by Step  and synthetic associated dataset (do not reuse without permission)
Data registered using AutoFinder Source Simulated Spot LM Parameters:

Source and target: ConvertBinarytoRoi 2 microns

AutoFinder Parameters:

Find small part in bigger field of view
; 10 microns; 50%

You need to first have identified ROI on both images, only their center will be considered.
The plugin Spot detector is a good choice here,
just make sure to have activated Export to Roi as output;
or eC-CLEM tool ConvertBinaryToPointRoi
IMPORTANT: Check your metadata first as it will be used by MyAutoFinder.

    • About the same content in both n-D images :
      This option will try to fit the full content of the image, assuming you have similar detections
    • Find Small Part in Bigger View (reverse or not) :
      The purpose here is to find an image position (typically EM)
      on a larger field of view (typically LM). The prealignment will be different
    • Max Error in microns: :
      A pointshould have a distance to its closest matching point below this value
      in order not to be considered as an outlier. Increase if no transformation was found.
      Rule of Thumb: about 10 pixels
    • Percentage of target point to keep: :
      This is the minimum percentage of point that have to match: 90% means almost no outliers
      50% or less if the number of detection are very different. 70% is usually a good trade off

Here is some example of semi automatic registration where lines in common are drawn to help the registration.

 

A wizard is also available

wizard

 

Deposit Digital Number: IDDN.FR.001.230005.000.S.P.2017.000.31500

This resource needs

Resources needing this

Leave a Review

Leave a review
Cancel review
View full changelog
Close changelog

Changelog

  • Version 1.0.0.3 • Released on: 2018-10-30 13:16:56
    Download
    Description:

    - added a faster mode for prealigned data
    - added time indication displayed in the output console

  • Version 1.0.0.2 • Released on: 2017-08-18 11:24:17
    Download
    Description:

    - fixed a bug on checkbox "export to ec-clem" for preparing data for further registration or analysis in EC-CLEM

  • Version 1.0.0.1 • Released on: 2017-04-11 12:25:34
    Download
    Description:

    - use new 3D ROI icy
    - Message can be removed (do not show again)

  • Version 1.0.0.0 • Released on: 2017-03-24 18:41:08
    Download
    Description:

    -- Add a new option to go on the registration with ECCLEM, allowing to
    then performed NONRIGID transformation from the automatically paitred
    points, or to generate error maps.

    -- SAVE paired and matching Rois in csv files , in installation directory of icy

    -- remove the limitation to 100 um of maximum distance

  • Version 0.0.2.1 • Released on: 2016-12-13 13:44:10
    Download
    Description:

    - propose automatically to set the maximum distance to about 10 pixels ( biggest pixels), expressed in micrometers
    --remove the limit of 10 um for maximum distance

  • Version 0.0.2.0 • Released on: 2016-11-30 05:42:57
    Download
    Description:

    - Modify ConvertbinarytoRoi so it can take Rois as well as inut (in addition to binary images), plus allow a resampling at the desired resolution.
    - Autofinder: changed the ransac procedure such that sampling are unique + distance-based.
    - bug correction on the find candidates
    - set few rules to avoid too few points to be matched.

  • Version 0.0.1.6 • Released on: 2016-11-08 11:53:31
    Download
    Description:

    - Add an interactive metadatacheck (as it is very important for autofinder)
    - Add the version number at the beginning

  • Version 0.0.1.5 • Released on: 2016-10-16 22:47:03
    Download
    Description:

    -correct typos.

  • Version 0.0.1.4 • Released on: 2016-10-02 11:54:19
    Download
    Description:

    -corrected an inversion in the wizard

  • Version 0.0.1.3 • Released on: 2016-10-01 18:05:16
    Download
    Description:

    - made the computatio more robust
    - Add a wizard tool

  • Version 0.0.1.2 • Released on: 2016-08-08 11:06:57
    Download
    Description:

    - Made it compliant with Java 1.6 >.
    - Added 2 parameters to adjust registration, also in block for protocols.
    - Corrected the matching process

  • Version 0.0.1.1 • Released on: 2016-08-04 15:20:01
    Download
    Description:

    non beta

  • Version 0.0.1.0b • Released on: 2016-08-04 15:04:26
    Download