Adaptive Gradient Balancing for Undersampled MRI Reconstruction and Image-to-Image Translation

Itzik Malkiel, Sangtae Ahn, Valentina Taviani, Anne Menini, Lior Wolf, Christopher J. Hardy

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Recent accelerated MRI reconstruction models have used Deep Neural Networks (DNNs) to reconstruct relatively high-quality images from highly undersampled k-space data, enabling much faster MRI scanning. However, these techniques sometimes struggle to reconstruct sharp images that preserve fine detail while maintaining a natural appearance. In this work, we enhance the image quality by using a Conditional Wasserstein Generative Adversarial Network combined with a novel Adaptive Gradient Balancing (AGB) technique that automates the process of combining the adversarial and pixel-wise terms and streamlines hyperparameter tuning. In addition, we introduce a Densely Connected Iterative Network, which is an undersampled MRI reconstruction network that utilizes dense connections. In MRI, our method minimizes artifacts, while maintaining a high-quality reconstruction that produces sharper images than other techniques. To demonstrate the general nature of our method, it is further evaluated on a battery of image-to-image translation experiments, demonstrating an ability to recover from sub-optimal weighting in multi-term adversarial training.

Original languageEnglish
Title of host publication2021 IEEE International Conference on Computational Photography, ICCP 2021
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781665419529
DOIs
StatePublished - 23 May 2021
Event13th IEEE International Conference on Computational Photography, ICCP 2021 - Haifa, Israel
Duration: 23 May 202125 May 2021

Publication series

Name2021 IEEE International Conference on Computational Photography, ICCP 2021

Conference

Conference13th IEEE International Conference on Computational Photography, ICCP 2021
Country/TerritoryIsrael
CityHaifa
Period23/05/2125/05/21

Keywords

  • MRI
  • WGAN
  • conditional generation
  • undersampled reconstruction

Fingerprint

Dive into the research topics of 'Adaptive Gradient Balancing for Undersampled MRI Reconstruction and Image-to-Image Translation'. Together they form a unique fingerprint.

Cite this