Dual-Encoder-Unet For Fast Mri Reconstruction
- Deep learning has shown great promise for successful acceleration of MRI data acquisition. A variety Deep Learning architectures have been proposed to obtain high fidelity image from partially observed kspace or undersampled image .
- U-Net has demonstrated impressive performance for providing high quality reconstruction from undersampled image data. The recently proposed dAutomap is an innovative approach to directly learn the domain transformation from source kspace to target image domain.
- However these networks operate only on a single domain where information from the excluded domain is not utilized for reconstruction. This paper provides a deep learning based strategy by simultaneously optimizing both the raw kspace data and undersampled image data for reconstruction.
Comparing Accelerated MR Reconstruction Models and Assessing Their Genereralizability to Datasets Collected with Different Coils
- The 2020 Multi-channel Magnetic Resonance Reconstruction (MC-MRRec)
Challenge had two primary goals: 1) compare different MR image reconstruction models on a large dataset and 2) assess the generalizability of these
models to datasets acquired with a different number of receiver coils (i.e., multiple channels). .
- This work provides relevant comparative information on the current state-of-the-art of MR reconstruction and highlights
the challenges of obtaining generalizable models that are required prior to
clinical adoption.
“We've always defined ourselves by the ability to overcome the impossible. And we count these moments.
These moments when we dare to aim higher, to break barriers, to reach for the stars, to make the unknown known
”
— Cooper, Interstellar
