Prior-Guided Image Reconstruction for Accelerated Multi-Contrast MRI via Generative Adversarial Networks
importance 5/10
partially solved
Advanced MRI Techniques and Applications medical imaging stage: research
Problem
Multi-contrast MRI scans take excessive time because each added contrast lengthens acquisition, limiting clinical use.
Why it matters
Long scan times reduce patient throughput and access to diagnostic information.
What the AI did
A conditional generative adversarial network synergistically recovers undersampled multi-contrast MRI acquisitions, combining reconstruction and synthesis to mitigate each framework's limitations.
Results
not stated in source
Impact
Peer-reviewed method that, per abstract, mitigates reconstruction limits; quantitative outcomes and adoption not stated.
Record
| Venue | IEEE Journal of Selected Topics in Signal Processing |
|---|---|
| Published | 2020 |
| Work type | article |
| Authors | Salman Ul Hassan Dar, Mahmut Yurt, Mo Shahdloo, M. Emrullah Ildiz, Berk Tınaz, Tolga Çukur |
| Citations | 174 (OpenAlex, at import time) |
| References | 113 |
| Model or method | conditional generative adversarial networks |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1109/jstsp.2020.3001737 |
| Open access | green |
| Keywords | MRI, GAN, image reconstruction, multi-contrast, deep learning |
Importance rating
5
real but narrower advance
Abstract states excessive multi-contrast MRI scan times limit clinical use and diagnostic access; real but narrower advance in scan efficiency, no scale stated.