NA-MIC Project Weeks

Back to Projects List

T1w MRI Artifact Removal

Key Investigators

• Owen Borders (Massachusetts General Hospital, USA)
• Sylvain Bouix (ETS, Canada)
• Charles Bricout (ETS, Canada)

Presenter location: In-person

Project Description

I will train a Generative Convolutional Neural Network to remove artifacts (motion, ghosting, metal artifacts, etc.) from T1w MRI images.

Objective

1. In the end, I plan to have a trained deep learning model and statistics on its performance.

Approach and Plan

1. I will write a Python program to generate artificial artifacts on TW1 MRI images. I will then train a CNN with the images with artificial artifacts as the input, and the original images without the artifacts as the output. Its performance will be evaluated on real data.

Progress and Next Steps

No response

Illustrations

No response

Background and References

No response