Dynamic imaging using motion-compensated smoothness regularization on manifolds (MoCo-SToRM)

@article{Zou2022DynamicIU,
  title={Dynamic imaging using motion-compensated smoothness regularization on manifolds (MoCo-SToRM)},
  author={Qing Zou and Luis A Torres and Sean B. Fain and Nara S. Higano and Alister J. Bates and Mathews Jacob},
  journal={Physics in Medicine \& Biology},
  year={2022},
  volume={67}
}
Objective. We introduce an unsupervised motion-compensated reconstruction scheme for high-resolution free-breathing pulmonary magnetic resonance imaging. Approach. We model the image frames in the time series as the deformed version of the 3D template image volume. We assume the deformation maps to be points on a smooth manifold in high-dimensional space. Specifically, we model the deformation map at each time instant as the output of a CNN-based generator that has the same weight for all time… 

References

SHOWING 1-10 OF 43 REFERENCES

Time-Dependent Deep Image Prior for Dynamic MRI

A generalized version of the deep-image-prior approach, which optimizes the weights of a reconstruction network to fit a sequence of sparsely acquired dynamic MRI measurements to reconstruct the continuous variation of dynamic MRI sequences with high spatial resolution.

Dynamic Imaging Using a Deep Generative SToRM (Gen-SToRM) Model

A generative smoothness regularization on manifolds (SToRM) model for the recovery of dynamic image data from highly undersampled measurements and introduces an efficient progressive training-in-time approach and an approximate cost function.

Deep Generative Storm Model for Dynamic Imaging

A novel generative smoothness regularization on manifolds (SToRM) model for the recovery of dynamic image data from highly undersampled measurements that reduces the memory demand and introduces efficient progressive approaches to minimize the computational complexity of the algorithm.

Motion correction of respiratory-gated PET images using deep learning based image registration framework

The proposed unsupervised non-rigid image registration framework using deep learning for motion correction uses a differentiable spatial transformer layer to warp the moving image to the fixed image and uses a stacked structure for deformation field refinement.

Free-Breathing and Ungated Dynamic MRI Using Navigator-Less Spiral SToRM

A kernel low-rank algorithm to recover free-breathing and ungated dynamic MRI from spiral acquisitions without explicit k-space navigators is introduced and provides improved reconstructions compared to the classical methods such as low- rank and XD-GRASP.

Free-breathing, zero-TE MR lung imaging

Three-dimensional radial, zero-echo time (TE) imaging for high-resolution, free-breathing magnetic resonance (MR) lung imaging using prospective and retrospective motion correction and Zero-TE appears to be an attractive pulse sequence for 3D isotropic lung imaging.

A Kernel-Based Low-Rank (KLR) Model for Low-Dimensional Manifold Recovery in Highly Accelerated Dynamic MRI

A novel algorithm with a kernel-based low-rank model generalizing the conventional low rank formulation is developed, which surpasses the conventionalLow rank-modeled approaches for dMRI.

Stable and efficient retrospective 4D-MRI using non-uniformly distributed quasi-random numbers

A non-uniform quasi random (NU-QR) reordering of the phase encoding lines was incorporated into 3D Cartesian acquisition and enables for stable 4D-MRI with high temporal and spatial resolution within short scan time for visualization of organ or tumor motion during free breathing.

Feasibility of free breathing Lung MRI for Radiotherapy using non-Cartesian k-space acquisition schemes.

The two non-Cartesian acquisition sequences, StarV IBE and SpiralVIBE, provide a free-breathing imaging protocol of the lung with sufficient image quality to permit anatomical, structural and functional assessment during radiotherapy.

Adam: A Method for Stochastic Optimization

This work introduces Adam, an algorithm for first-order gradient-based optimization of stochastic objective functions, based on adaptive estimates of lower-order moments, and provides a regret bound on the convergence rate that is comparable to the best known results under the online convex optimization framework.