MotionDPS is a unified Bayesian framework for motion-compensated 3D MRI reconstruction that alternates diffusion posterior updates with proximal optimization for rigid motion and coil sensitivity estimation using pretrained 3D complex-valued score-based diffusion models as anatomical priors.
Deep learning for accelerated and robust MRI reconstruction,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
MotionDPS: Motion-Compensated 3D Brain MRI Reconstruction
MotionDPS is a unified Bayesian framework for motion-compensated 3D MRI reconstruction that alternates diffusion posterior updates with proximal optimization for rigid motion and coil sensitivity estimation using pretrained 3D complex-valued score-based diffusion models as anatomical priors.