Project-MONAI / Project-MONAI/MONAI
Transform for noise for MRI
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 8.7k
- Forks
- 1.6k
- Avg merge
- 5d 1h
- Merged PRs (30d)
- 20
Description
Hello, would it be possible to add noise for MRI? Specifically, modern MRI uses multiple coils which results in a non-central chi distribution for the noise (https://en.wikipedia.org/wiki/Noncentral_chi_distribution), which is a generalization of Rician noise (i.e., for a single quadrature coil they're the same). Today, nearly all (maybe all?) human scans are received with multiple coils. Probably all large brain datasets are received with multiple coils. For brain, for example, 32-64 is the standard today. Thanks very much for your great package!
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by reviewing MONAI's existing noise transforms and their tests to understand the transform API and parameter conventions. Then clarify the noncentral chi distribution parameters, expected MRI input behavior, and test cases; done means a documented transform with coverage for the proposed multi-coil noise model.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python, pytorch
- Domain
- computer-vision
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100