Welcome to our Github page dedicated to Random Illumination Microscopy images reconstruction!
This page is linked to the following publication:
Mangeat et al., Super-resolved live-cell imaging using random illumination microscopy, Cell Reports Methods (2021),
https://doi.org/10.1016/j.crmeth.2021.100009
AlgoRIM is our image reconstruction interface allowing to achieve super-resolution based on the properties of the Random Illumination Microscopy (RIM). Download AlgoRIM and test it! You will find below a tutorial to guide you step by step.
If you want the images shown in our publication please send an email to thomas.mangeat@univ-tlse3.fr or simon.labouesse@univ-tlse3.fr
- You are free to use this software for research purposes only, but you must not transmit and distribute it without our consent (See contacts list).
- In addition, you undertake to include a citation, Mangeat et al., Super-resolved live-cell imaging using random illumination microscopy, Cell Reports Methods (2021), https://doi.org/10.1016/j.crmeth.2021.100009, whenever you present or publish results that are based on it.
- The CNRS makes no warranties of any kind on this software and shall in no event be liable for damages of any kind in connection with the use and exploitation of this technology.
Latest version (07/01/22) AlgoRIM V1.2
https://drive.google.com/file/d/1-5DktHl8sr4dykBrrwYh86hwuHcNbJF0/view?usp=sharing
Release notes :
- Better denoising and data normalization
- Less iterations: algorithm converges faster
- Internet connection
- Administrator rights on the computer (In some instances the antivirus try to prevent the installation)
- Windows operating system (Tested on : Windows 10 and 11 pro)
- Use this link and download all the content: https://drive.google.com/file/d/1-5DktHl8sr4dykBrrwYh86hwuHcNbJF0/view?usp=sharing
- Unzip the folder.
- You will find a folder named "data" that contains the file we want to reconstruct for this example and the PSF we need.
- Start the download of AlgoRIM by clicking on "AlgoRIM_V1_2_installer_web.exe". Follow the process. This operation may take several minutes. Don't forget to add a shorcut to your desktop.
- Once your installation is complete, open AlgoRIM_V1_2 interface.
- Images from: 'Other'
- Input data: 'Only 1 file (Stream)'
- File reconstruction mode: 'Full file'
- Select: Go in the folder you previously unziped, then in he folder 'data' select 'ROI1_SUM_image_43_800.tif'
- PSF emission: 'PSF256_520nm.tif'
- PSF excitation: 'PSF256_405nm.tif'
- Image expansion factor: 2
- PSF expansion factor: 1
- Regularization parameter: 0.15
- Number of iterations: 4
- Pre-filtering parameter: 0.05. You can click on the 'Adjust' button to adjust it.
- Make sure that the cutoff frequency matches your system.
- You can click on 'Launch preview' to start preview. Use the 'STOP' button if you don't want to wait until the end of the iterations.
- Click on 'Launch' to reconstruct the file.
- This will create a reconstruction folder in 'data' folder, with the reconstructed image in it.
Allow the interface to be able to identify and differentiate your channels.
- If you select Other: No prerequisite on the name of the channels but they will not be differentiated during reconstruction. You can create a subfolder for each channel and reconstructed them separately.
- Inscoper: channel 0 is identify with 'c0' in the image name and channel 1 with 'c1'
- Abbelight: the image name begins with 'ROI2' or 'ROI1', depending on the channel
Select the type of the input data
- Global folder: Launch several reconstructions in a row by selecting a folder containing several projects.
It is based on Inscoper's raw image backup mode: global_folder/sub_folder/images/RAW_DATA/. If you want you launch preview, you need to select precisely the raw data folder you want to launch preview on.- Raw data folder: Launch reconstruction on one folder containning .tif files.
- Only 1 file (Stream): With this mode you can launch reconstruction on one .tif file.
Select how to reconstruct the input data
- Full file: 1 raw data file => 1 reconstructed file
- Sequential: Set the number of raw images per reconstructed image.
- Interleaved: Like the Sequential mode, it divides file in sequences but they are interleaved. The 1st speckle of the n+1 sequence is x speckles after the 1st speckle of the n sequence. x is the gap value. It improves time resolution and denoising.
- Folder must contain raw data files of one or more raw data image.
- Files must be in .tif format.
Over-sampling factor:
Raw images are usually recorded with a pixel-size about lambda/(4xNA), with lambda the wavelength of the emitted light and NA the numerical aperture of the objective. The super-resolved reconstruction should be discretized with a smaller pixel size. The over-sampling factor is the ratio between the pixel size of the raw images and the pixel size of the reconstructed (super-resolved) image, e.g., 2 will produce an output/reconstructed image with a pixel size that is half the native pixel size of the raw data.Remark: the reconstruction code is based on Fast Fourier Transforms, which are faster if the number of pixels along any direction is a power of 2 (128, 256, 512, etc.). Of course this is not mandatory. Images must be square.
Collection PSF:
PSF of the objective with respect to the collected light (emitted by the fluorophores). This PSF can be estimated experimentally or generated theoretically with a free software like 'PSF generator'.The PSF file should be provided in .tif format and the maximum of the PSF should be at the center of the image. Beware that this image should be provided with a pixel size corresponding to that of the super-resolved/reconstructed image.
Excitation PSF:
PSF of the objective with respect to the illlumination light. This PSF corresponds to the autocorrelation of the speckle; it can be generated with a free software like 'PSF generator'. Note that this PSF is not affected by any aberration.The PSF file should be provided in .tif format and the maximum of the PSF should be at the center of the image. This image should be provided with a pixel size corresponding to that of the super-resolved/reconstructed image.
Number of iterations:
Number of iterative updates in the variance matching algorithm. The early stopping of the iterative scheme is producing a Tikhonov-like regularization of the reconstruction. You can adjust this setting with the preview mode showing the progression of the reconstructed image over the course of the algorithm.Pre-filtering parameter:
Each raw image is deconvolved using a Wiener filtering. You can adjust this parameter with the adjust button in the preview mode.This mode allows you to preview the reconstruction result on the first raw image. If you want to interrupt it, please use the stop button.
This mode starts the reconstruction and saves your results in a folder named according to your settings.
Distribution request:
- Simon Labouesse (AlgoRIM development and implementation)
- Claire Estibal (AlgoRIM interface development)
- Thomas Mangeat (Methodological part for biological reconstruction and imaging workflow)
- Anne Sentenac
- Jérôme Idier
- Marc Allain