Learning Task-Specific Strategies for Accelerated MRI

Abstract

Compressed sensing magnetic resonance imaging (CS-MRI) seeks to recover visual information from subsampled measurements for diagnostic tasks. Traditional CS-MRI methods often separately address measurement subsampling, image reconstruction, and task prediction, resulting in a suboptimal end-to-end performance. In this work, we propose TACKLE as a unified co-design framework for jointly optimizing subsampling, reconstruction, and prediction strategies for the performance on downstream tasks. The naïve approach of simply appending a task prediction module and training with a task-specific loss leads to suboptimal downstream performance. Instead, we develop a training procedure where a backbone architecture is first trained for a generic pre-training task (image reconstruction in our case), and then fine-tuned for different downstream tasks with a prediction head. Experimental results on multiple public MRI datasets show that TACKLE achieves an improved performance on various tasks over traditional CS-MRI methods. We also demonstrate that TACKLE is robust to distribution shifts by showing that it generalizes to a new dataset we experimentally collected using different acquisition setups from the training data. Without additional fine-tuning, TACKLE leads to both numerical and visual improvements compared to existing baselines. We have further implemented a learned 4$\times$-accelerated sequence on a Siemens 3T MRI Skyra scanner. Compared to the fully-sampling scan that takes 335 seconds, our optimized sequence only takes 84 seconds, achieving a four-fold time reduction as desired, while maintaining high performance.

Our code is available at https://github.com/zihuiwu/TACKLE. More information can be found at http://imaging.cms.caltech.edu/tackle/

References in corpus (18)

• VoxelMorph: A Learning Framework for Deformable Medical Image Registration
• Estimating or Propagating Gradients Through Stochastic Neurons for Conditional Computation
• MoDL: Model Based Deep Learning Architecture for Inverse Problems
• Learned Primal-dual Reconstruction
• Compressed Sensing MRI Reconstruction using a Generative Adversarial Network with a Cyclic Loss
• BM3D Frames and Variational Image Deblurring
• fastMRI: An Open Dataset and Benchmarks for Accelerated MRI
• Self-Supervised Learning of Physics-Guided Reconstruction Neural Networks without Fully-Sampled Reference Data
• Plug and play methods for magnetic resonance imaging (long version)
• Plug-and-Play Methods for Integrating Physical and Learned Models in Computational Imaging
• RARE: Image Reconstruction using Deep Priors Learned without Ground Truth
• J-MoDL: Joint Model-Based Deep Learning for Optimized Sampling and Reconstruction
• Dense Recurrent Neural Networks for Accelerated MRI: History-Cognizant Unrolling of Optimization Algorithms
• B-spline Parameterized Joint Optimization of Reconstruction and K-space Trajectories (BJORK) for Accelerated 2D MRI
• SGD-Net: Efficient Model-Based Deep Learning with Theoretical Guarantees
• JoJoNet: Joint-contrast and Joint-sampling-and-reconstruction Network for Multi-contrast MRI
• Stochastic Optimization of 3D Non-Cartesian Sampling Trajectory (SNOPY)
• CoRRECT: A Deep Unfolding Framework for Motion-Corrected Quantitative R2* Mapping