Department of Biomedical Magnetic Resonance UKT and High-Field MR Center MPI
Our goal is the development and application of novel magnetic resonance techniques at very high magnetic fields to specifically probe the anatomical and functional microstructure of the brain. We aim to understand how physiological processes and microstructure impact the measured resonance signal, and how these magnetic fingerprints can be used to reliably detect brain activation.
Current Projects
Imaging speed is a key factor to capture rapid changes at high spatial and temporal resolution. A major limitation of magnetic resonance (MR) imaging is its rather low speed compared to other modalities like ultra sound or computerized tomography. We aim to explore two novel concepts to boost MR imaging speed by another order of magnitude compared to existing techniques. SpreadMRI fundamentally steps beyond current concepts of image encoding by exploiting a spectral spin modulation that so far has not been utilized. SpreadMRI is based on the rapid and local modulation of magnetic fields produced by current loops and/or radiofrequency (RF) loops. Applied spectral modulations are in the MHz range bridging the low-frequency band of switched gradients (kHz) and the 100 MHz range of the Larmor frequency. SpreadMRI spreads the bandwidth of gradient-encoded spin frequencies using distinct carrier frequencies originating from a certain region of the object. This spatially unique information will then be used to disentangle parts of the object, and thus to drastically boost imaging speed.
These three cartoons aim to visualize the rapid steering of local magnitude receive coil profiles during the acquisition process. This increases the spatial information content compared to static fields as use in conventional parallel imaging.
Modeling of Microvascular Effects in Cerebral Cortical Laminar fMRI
What are the limitations of fMRI and how faithfully can it reflect neural activity? In a three-year US-German project, we will improve our ability to measure neural activity from fMRI through detailed biophysical modelling of the hemodynamic response and corresponding high-resolution MRI measurements. The group of Jon Polimeni at MGH is currently developing the framework for generating realistic microvascular networks and dynamics to simulate hemodynamic responses to neural activity. These networks will allow us to simulate different fMRI contrasts, neuronal activation durations, cortical orientations and vascular densities. The artificial vessel models will then be combined with real macrovascular architectures derived from very high-resolution MR images and capillary densities estimated from cerebral blood volume (CBV) measurements performed at the 9.4 T in Tübingen. To achieve high robustness against subject motion, a camera-based prospective motion correction system will be used.
Synthetic BOLD predictions on a microvascular reconstruction in mouse brain.
UHF-NeuroBOOST
Main objectives of the UHF-NeuroBoost project are the developments of the novel pTx optimized 16-element Tx/64-element Rx RF array coils with extended transmit (Tx) and receive (Rx) coverage for human head/neck UHF neuro-MRI at 7T, 9.4T and 11.7T. The optimization of the array will be carried out with an evaluation of the pTx pulse performance guiding the design process. We will optimize both the Tx performance and SNR of the arrays. Since at UHF, dipole/loop combined arrays have been shown to improve both the Tx performance and central SNR as compared to common loop arrays, we will consider combining loops and dipoles in designing the arrays. As part of the development, we will compare a common double-layer ToRo setup, which consists of two nested Tx and Rx arrays, with a novel hybrid type of single-layer RF array coil consisting of a combination of TxRx elements with Rx-only elements. We also plan incorporating local B0 shim loops into the RF coil structure.
Project overview. Optimization of the transmit (WP2) and receive (WP3) performance, which are tightly connected with the development of pulse sequences (WP1), will be followed by the development of the B0 shimming setup (WP4) and anatomical housing (WP5). Three UHF arrays will be constructed in WP6. After constructing (WP6), safety evaluation (WP7), all array coils will be tested in-vivo and compared with common array designs (WP8).
Probing the Brain Using Long-Term Implantable e-Skin Magnetic Resonance Imaging Sensors (E-Brain BW-Foundation)
Thanks to its unique specificity and the absence of ionizing radiation, magnetic resonance imaging (MRI) is one of the prime modalities to assess brain function and metabolism. However, the spatial and temporal resolution of MRI is still limited compared to optical or electrophysiological methods. Even at very high magnetic fields up to 10 Tesla or more, it is highly challenging to acquire signals not only reflecting gross activation or metabolic changes of larger cortical areas but to probe functional cortical subunits such as layers or columns within the cerebral cortex. The two main reasons for this limitation are the low intrinsic sensitivity of MR, and the spatially unspecific coupling between neuronal excitation, local blood regulation, and energy metabolism. Therefore, the major objective of the proposed E-Brain project is to overcome these fundamental limitations of conventional MRI by applying an innovative MR detection concept that combines a highly sensitive integrated circuit-based readout with fully biocompatible, long-term implantable microcoil arrays. These ultra-thin coil arrays with a thickness of only a few micrometers, which produce virtually no tissue damage during implantation, can pick up and amplify the MR signal with unprecedented sensitivity and local specificity.
