Typically, these are almost acquisitions of lower quality (eg wrt resolution, contrast, or artefacts) to slower scans with higher resolution; At the sametime we know that the object is primarily left apart from potential non-linear deformations and the presence of an interventional tool (eg a needle).
Consequently, a lot is known about the object.
This project aims to include available prior knowledge, and perfrom undersampled MR reconstruction using Deep Learning.
Tel.:+49 391 6117119