algorithm_developer_task

technical task for potential seetreat algorithm developers

Task

• Write a python code which
  • registers on CT image to another CT image using deformable image registration
  • Quantifies and visualised the deformable vector field (DVF)

You may use any third party libraries you wish.

Please implement your solution in a private repository, and add brendan-whelan-seetreat and lars-mejnertsen-seetreat as collaborators.

Data

The data you will use in this excecise can be downloaded here:

https://zenodo.org/records/7260705

Specifically:

• Download Task1.zip (~nb: this is 15 Gb so allow some time for download)
• Register pelvis/2PA001/ct.nii.gz (the moving image) to pelvis/2PA002/2PA002 (the static image)

Specifically, register

Further information

• We don't ask you to spend more than ~2 hours on this. We don't really care if you finish it or not, we are more interested in how you think and communicate. It's totally fine to just tell us what you would have done had you had more time.
• We are looking for:
  • Working code
  • Well written code
  • Ability to verbally explain the underlying algorithms or libraries which are utilised
  • Applicability of the solution to datasets other then the development dataset provided above