Accelerated Flow and Motion MRI

Magnetic resonance imaging (MRI) facilitates the acquisition of images of the inside of the human body and this has made MRI an indispensable tool in modern healthcare. Apart from providing anatomical contrast, MRI has numerous other ways to distinguish normal from pathological anatomy, function, or even metabolism. One of these ways is 4D flow MRI, which is a specialized MRI scan technique that enables the quantification of tissue motion/velocity in a 3D and time-resolved manner.
Conventional 3D velocity measurements require huge amounts of data, intrinsic to the multi-dimensional nature of the acquisition, which generally results in unfavorable long scan times. Long scan times can be shortened at the expense of lower spatiotemporal resolution or by undersampling of the MRI scan. The latter requires a suitable reconstruction technique to correct image artifacts resulting from undersampling.
In this thesis, a novel technique for accelerated flow and motion MRI based on incoherent undersampling of the data acquisition in combination with compressed sensing reconstruction is introduced. This innovation in accelerated 4D flow MRI has been validated and compared to existing techniques. The accuracy of the developed technique was quantified using velocity, flow, and derived hemodynamic parameters such as wall shear stress as main outcome parameters. Moreover, the technique now has been implemented for clinical use in our local medical center and the required steps towards clinical implementation have been made available to the wider radiological community.