Premium
Systematic evaluation of iterative deep neural networks for fast parallel MRI reconstruction with sensitivity‐weighted coil combination
Author(s) -
Hammernik Kerstin,
Schlemper Jo,
Qin Chen,
Duan Jinming,
Summers Ronald M.,
Rueckert Daniel
Publication year - 2021
Publication title -
magnetic resonance in medicine
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.696
H-Index - 225
eISSN - 1522-2594
pISSN - 0740-3194
DOI - 10.1002/mrm.28827
Subject(s) - computer science , regularization (linguistics) , artificial intelligence , consistency (knowledge bases) , robustness (evolution) , artificial neural network , iterative reconstruction , machine learning , deep learning , test data , sensitivity (control systems) , data consistency , pattern recognition (psychology) , biochemistry , chemistry , gene , programming language , operating system , electronic engineering , engineering
Purpose To systematically investigate the influence of various data consistency layers and regularization networks with respect to variations in the training and test data domain, for sensitivity‐encoded accelerated parallel MR image reconstruction. Theory and Methods Magnetic resonance (MR) image reconstruction is formulated as a learned unrolled optimization scheme with a down‐up network as regularization and varying data consistency layers. The proposed networks are compared to other state‐of‐the‐art approaches on the publicly available fastMRI knee and neuro dataset and tested for stability across different training configurations regarding anatomy and number of training samples. Results Data consistency layers and expressive regularization networks, such as the proposed down‐up networks, form the cornerstone for robust MR image reconstruction. Physics‐based reconstruction networks outperform post‐processing methods substantially for R = 4 in all cases and for R = 8 when the training and test data are aligned. At R = 8, aligning training and test data is more important than architectural choices. Conclusion In this work, we study how dataset sizes affect single‐anatomy and cross‐anatomy training of neural networks for MRI reconstruction. The study provides insights into the robustness, properties, and acceleration limits of state‐of‐the‐art networks, and our proposed down‐up networks. These key insights provide essential aspects to successfully translate learning‐based MRI reconstruction to clinical practice, where we are confronted with limited datasets and various imaged anatomies.