fastMRI: An Open Dataset and Benchmarks for Accelerated MRI
This work provides an open dataset and benchmarks to help the medical imaging and machine learning communities advance MR image reconstruction, though it is incremental in nature as it builds on existing efforts by standardizing evaluation.
The authors tackled the problem of accelerating MRI by introducing the fastMRI dataset, a large-scale collection of raw MR measurements and clinical images, along with standardized benchmarks, to facilitate machine-learning approaches for MR image reconstruction, aiming to reduce medical costs and patient stress.
Accelerating Magnetic Resonance Imaging (MRI) by taking fewer measurements has the potential to reduce medical costs, minimize stress to patients and make MRI possible in applications where it is currently prohibitively slow or expensive. We introduce the fastMRI dataset, a large-scale collection of both raw MR measurements and clinical MR images, that can be used for training and evaluation of machine-learning approaches to MR image reconstruction. By introducing standardized evaluation criteria and a freely-accessible dataset, our goal is to help the community make rapid advances in the state of the art for MR image reconstruction. We also provide a self-contained introduction to MRI for machine learning researchers with no medical imaging background.