Memory-efficient Learning for High-Dimensional MRI Reconstruction

TitleMemory-efficient Learning for High-Dimensional MRI Reconstruction
Publication TypeConference Paper
Year of Publication2021
AuthorsWang, K., Kellman M., Sandino C. M., Zhang K., Vasanawala S. S., Tamir J. I., Yu S. X., & Lustig M.
Published inProceedings of International Society for Magnetic Resonance in Medicine
Date Published05/2021
Other NumbersarXiv:2103.04003
Keywordsmagnetic resonance imaging (MRI), memory-efficient learning, unrolled reconstruction

Deep learning (DL) based unrolled reconstructions have shown state-of-the-art performance for under-sampled magnetic resonance imaging (MRI). Similar to compressed sensing, DL can leverage high-dimensional data (e.g. 3D, 2D+time, 3D+time) to further improve performance. However, network size and depth are currently limited by the GPU memory required for backpropagation. Here we use a memory-efficient learning (MEL) framework which favorably trades off storage with a manageable increase in computation during training. Using MEL with multi-dimensional data, we demonstrate improved image reconstruction performance for in-vivo 3D MRI and 2D+time cardiac cine MRI. MEL uses far less GPU memory while marginally increasing the training time, which enables new applications of DL to high-dimensional MRI.


Magna Laude Award