MR Physics and Signal Processing
MR pulse sequence design
The interplay of radio-frequency excitation, magnetic field gradients and tissue properties offers a wealth of image contrasts in Magnetic Resonance (MR) imaging and spectroscopy. Our research focuses on developing MR pulse sequences to achieve optimal contrast- and signal-to-noise ratios in the shortest possible scan time. Moreover, we explore novel contrast mechanisms beyond the currently available set of anatomical and functional imaging contrasts for various applications. To this end, we employ computer simulations and experimental imaging to design and validate novel MR sequence schemes and methods.
Reduced data acquisition concepts
The image encoding process in MR is based on a sequential spectroscopic principle. While this strategy offers great flexibility regarding the achievable resolution, it also implies a relatively slow data acquisition process. To speed up MR data acquisition, data may be undersampled. To resolve the resulting signal aliasing, the concept of signal detection with parallel receivers is used. In addition and beyond, properties of the object to be imaged may be exploited to reduce undersampling artifacts. We design, develop and implement undersampling strategies and appropriate image reconstruction techniques to enable highly accelerated imaging.
Signal modeling and correction
The MR signal is inherently weak and noise originating from thermal sources in the object being imaged as well as object motion and system imperfections can compromise image quality significantly. We develop and implement model-based methods to reduce image artifacts and noise to allow for recovery of physically and physiologically meaningful parameters of the imaged object. The work exploits prior and measured information of object properties including fundamental relations from MR physics, tissue and fluid mechanics. To this end, we devise signal modeling, correction and recovery approaches for robust and accurate quantitative imaging.