README.md

emboBioImage2023_registration

Materials for the advanced image registration section of the "Advanced methods in bioimage analysis" EMBO 2023 course

Course outline

1 Lecture

1. Watch the lecture video
2. (Optional) Download the lecture slides

2 Setup

If working on EMBL's BAND Platform, run these once:

• Open a terminal (click the icon near the top of the screen, or Applications > System tools > MATE terminal)
• Run /scratch/bioimagecourse2023/session4/get-data.sh

Now you should have a folder in your home directory called emboBioImage2023_registration. It should contain a folder called sampleImages containing six files.

When you're ready to start the hands-on portion run these two commands in a terminal.

• source /scratch/bioimagecourse2023/session4/setup.sh
• /scratch/bioimagecourse2023/session4/notebook.sh

Otherwise:

1. Check out this repository
2. Run ./downloadSampleImages to ... download more sample images
3. Set up your environment, if necessary (see below)

3 Practical / Hands-on

Each notebook contains instructions and comments. To get the most out of them, students should read and follow instructions in the notebooks, run the notebook cell-by-cell, ensuring they understand the code. Also important:

• Predict the output / result of each cell before running
• Observe the result after running the cell
• Compare to your prediction

Unsurprisingly, the recommended order for notebooks is:

1. 1_resample_images.ipnb
2. 2_translation_initialization.ipnb
3. 3_similarity_metrics.ipnb
4. 4_multi_resolution.ipnb
5. 5_nonlinear.ipynb

Dependencies

• simple-elastix
• itk-widgets

Registration and related tools

• elastix
• ANTs
• CMTK
• BigStream (for big data)
  • bigstream video tutorial
• bUnwarpJ (by Ignacio!)
• ec-Clem (by Perrine!)
• BigWarp (by John!)
  • A playlist of bigwarp videos
• BigStitcher
• ABBA
• Iterative closest point (ICP): classical algorithm with many implementations
• Coherent point drift
  • a copy of the original matlab code
  • a python reimplementation
• Janelia "render services"

Acknowledgements

Thank you to the organizers of the course: Ignacio Arganda-Carreras, Anna Klemm, Perrine Paul-Gilloteaux, Christian Tischer, Irena Provaznikova, and Alba Catherine Casas Saavedra. The Saalfeld lab and HHMI Janelia for supporting me, and all the course particpants.

Image credits

• CT / MR
  • Adapted from the 2007 MIT registration course by Zollei and Wells
• JFRC2010
  • Jenett et al 2012
• JFRC2018
  • Bogovic et al 2020