Active MR k-space Sampling with Reinforcement Learning

  title={Active MR k-space Sampling with Reinforcement Learning},
  author={L. Pineda and Sumana Basu and Adriana Romero and Roberto Calandra and Michal Drozdzal},
Deep learning approaches have recently shown great promise in accelerating magnetic resonance image (MRI) acquisition. The majority of existing work have focused on designing better reconstruction models given a pre-determined acquisition trajectory, ignoring the question of trajectory optimization. In this paper, we focus on learning acquisition trajectories given a fixed image reconstruction model. We formulate the problem as a sequential decision process and propose the use of reinforcement… 


This work generalizes DPS to a sequential method that actively picks the next sample based on the information acquired so far; dubbed Active-DPS (ADPS), and demonstrates strong performance in active acquisition Magnetic Resonance Image (MRI) reconstruction, outperforming DPS and other deep learning methods.

Uncertainty-Driven Adaptive Sampling via GANs

An adaptive sampling method for the linear model, driven by the uncertainty estimation with a generative adversarial network (GAN) model, is proposed, and it is empirically demonstrated that the posterior estimate contracts rapidly towards the correct mode.

Experimental design for MRI by greedy policy search

This work proposes to learn experimental design strategies for accelerated MRI with policy gradient methods, and empirically shows that adaptivity is key to improving subsampling designs.

Active Deep Probabilistic Subsampling

It is validated that A-DPS improves over DPS for MNIST classifcation at high subsampling rates and demonstrates strong performance in active acquisition Magnetic Resonance Image (MRI) reconstruction, outperforming DPS and other deep learning methods.

Deep reinforcement learning and its applications in medical imaging and radiation therapy: a survey

The basics of reinforcement learning are introduced and various categories of DRL algorithms and DRL models developed for medical image analysis and radiation treatment planning optimization are reviewed, which can resolve the challenges from scarce and heterogeneous annotated medical image data.

End-to-End Sequential Sampling and Reconstruction for MR Imaging

A fully differentiable framework that jointly learns a sequential sampling policy simultaneously with a reconstruction strategy is proposed that outperforms the current state-of-the-art learned κ-space sampling baseline on up to 96.96% of test samples.

Deep Reinforcement Learning for Beam Angle Optimization of Intensity-Modulated Radiation Therapy

The proposed DRL-based beam angle selection strategy can generate personalized beam angles within a few seconds, and the resulting treatment plan is superior to that obtained using evenly distributed angles.

Adaptive Sampling for Linear Sensing Systems via Langevin Dynamics

A Bayesian method for adaptive sampling based on greedy variance reduction and stochastic gradient Langevin dynamics (SGLD) is proposed to accelerate the acquisition of magnetic resonance imaging (MRI).

On the Feasibility of Machine Learning Augmented Magnetic Resonance for Point-of-Care Identification of Disease

The feasibility of an ML -augmented MR pipeline that directly infers the disease sidestepping the image reconstruction process is explored, and comparable performance to image-based classifiers, trained on images reconstructed with full k -space data, can be achieved using small quantities of data.

DSFormer: A Dual-domain Self-supervised Transformer for Accelerated Multi-contrast MRI Reconstruction

  • Bo ZhouJo Schlemper M. Sofka
  • Computer Science
    2023 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV)
  • 2023
A dual-domain self-supervised transformer (DSFormer) for accelerated MC-MRI reconstruction which generates high-fidelity reconstructions which outperform current fully-super supervised baselines and approach the performance of full supervision.



Accelerating magnetic resonance imaging via deep learning

This paper proposes a deep learning approach for accelerating magnetic resonance imaging (MRI) using a large number of existing high quality MR images as the training datasets. An off-line

Self-Supervised Deep Active Accelerated MRI

This work proposes to simultaneously learn to sample and reconstruct magnetic resonance images (MRI) to maximize the reconstruction quality given a limited sample budget, in a self-supervised setup by considering both the data acquisition and the reconstruction process within a single deep-learning framework.

A Deep Cascade of Convolutional Neural Networks for Dynamic MR Image Reconstruction

A framework for reconstructing dynamic sequences of 2-D cardiac magnetic resonance images from undersampled data using a deep cascade of convolutional neural networks (CNNs) to accelerate the data acquisition process is proposed and it is demonstrated that CNNs can learn spatio-temporal correlations efficiently by combining convolution and data sharing approaches.

Variable-Density Single-Shot Fast Spin-Echo MRI with Deep Learning Reconstruction by Using Variational Networks.

Compared with the conventional parallel imaging and compressed sensing reconstruction (PICS), the variational network (VN) approach accelerates the reconstruction of variable-density single-shot fast spin-echo sequences and achieves improved overall image quality with higher perceived signal-to-noise ratio and sharpness.

Reducing Uncertainty in Undersampled MRI Reconstruction With Active Acquisition

A novel method is presented that, at inference time, dynamically selects the measurements to take and iteratively refines the prediction in order to best reduce the reconstruction error and, thus, its uncertainty.

Learning-Based Compressive MRI

A learning-based framework for optimizing MRI subsampling patterns for a specific reconstruction rule and anatomy, considering both the noiseless and noisy settings is proposed and a novel parameter-free greedy mask selection method is presented.

fastMRI: An Open Dataset and Benchmarks for Accelerated MRI

The fastMRI dataset is introduced, 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.

Multi-channel Generative Adversarial Network for Parallel Magnetic Resonance Image Reconstruction in K-space

By experimentally comparing with other methods, it is demonstrated that multi-channel GAN can perform image reconstruction with an affordable computation cost and an imaging acceleration factor higher than the current clinical standard.

Recurrent inference machines for reconstructing heterogeneous MRI data