Abstract: A method for medical imaging performs a sparse-sampled tomographic imaging acquisition by an imaging system to produce acquired sparse imaging samples; synthesizes by a first deep learning network unacquired imaging samples from the acquired imaging samples to produce complete imaging samples comprising both the acquired imaging samples and unacquired imaging samples; transforms by a physics module the complete imaging samples to image space data based on physics and geometry priors of the imaging system; and performs image refinement by a second deep learning network to produce tomographic images from the image space data. The physics and geometry priors of the imaging system comprise geometric priors of a physical imaging model of the imaging system, and prior geometric relationships between the sample and image data domains.

Abstract: The following relates generally to motion prediction in magnetic resonance (MR) imaging. In some embodiments, a “modular” approach is taken to motion correction. That is, individual motion sources (e.g., a patient's breathing, heartbeat, stomach contractions, peristalsis, and so forth) are accounted for individually in the motion correction. In some embodiments, to correct for a particular motion source, a reference state is created from a volume of interest (VOI), and other states are created and deformably aligned to the reference state.