Machine learning techniques were first introduced six years ago to solve the inverse problem of MR image generation from accelerated acquisitions (1,2,3). Since then, the field has grown significantly, leading to a wide range of machine learning methods developments that can be applied to a wide range of imaging applications that have already been rolled out as clinical products by major manufacturers.

My presentation begins by first focusing on the background of an artificial intelligence process to generate MR images stemming from acquired measurement data. In particular, I will discuss a range of approaches that map iterative algorithms onto neural networks. I will discuss advantages and ongoing challenges, covering the design of neural network architecture and the training procedure, error metrics, computation time, generalizability and validation of the results.  I will also include a discussion of the lessons learned from the recent fastMRI image reconstruction challenges organized jointly with Facebook AI research (4,5).

References
1. Learning a variational model for compressed sensing MRI reconstruction. Hammernik, et al. Proc. ISMRM p33 (2016).
2. Accelerating magnetic resonance imaging via deep learning. Wang et al. IEEE ISBI 514-517 (2016).
3. Hammernik et al. Learning a Variational Network for Reconstruction of Accelerated MRI Data. MRM, 79:3055-3071 (2018).
4. Advancing machine learning for MR image reconstruction with an open competition: Overview of the 2019 fastMRI challenge. Knoll et al. MRM 84 (6), 3054-3070 (2020).
5. Results of the 2020 fastmri challenge for machine learning MR image reconstruction. Muckley et al. IEEE TMI 40 (9), 2306-2317(2021).