The aim of our research is to establish new methods for investigating the microscopic properties of tissue using its NMR characteristics.
MR-observable properties, for example the relaxation times of various nuclear species, are determined by different microscopic properties of the environment averaged over a distance sampled through diffusion of the nuclei during the NMR encoding time (~ 10 microns for brain tissue). The properties thus sampled and averaged include electric field gradients (for nuclei with a quadrupole moment), local magnetic field perturbations related to the distribution of different elements, motion restriction by boundaries, and concentration of specific substances (proteins, myelin, ferritin).
This is clearly a rich source of information, which has not been fully exploited until now, partly because of its complexity. For the human/mammal brain, the sensitivity of the MR relaxation times to tissue type (white matter/ grey matter) and its pathological changes (e.g. tumours) has been noticed more than 30 years ago, and the diagnostic powers of the MR contrast has revolutionized modern medicine. Learning more about the microscopic structure which gives rise to this contrast is our long-term aim.
One direction of research focuses on the measurement of different “traditional” NMR parameters (relaxation times, proton density) with high precision and accuracy using optimized methods/protocols and postprocessing methods.
A related direction concentrates on the development of methods for quantitative imaging at very high magnetic fields. Phase and susceptibility contrast and tissue T2* contrast in high resolution imaging at very high fields are already starting to provide a deeper insight into brain structure than achievable at lower fields. Correlation of these properties with the element distribution in tissue will help clarify the origin of the contrast.
Very high-resolution imaging is a brute-force tool to characterize tissue microscopy down to the diffusion limit (~ 10 microns). We have an ongoing program of investigating different objects with very high resolution quantitative imaging and/or optimised contrast. Furthermore, we investigate possibilities for exploiting the extensive averaging required by high-resolution imaging. As a first step, a method for reconstruction of thin slices from multiple acquisition of thick slices (superresolution reconstruction) has been developed.