Magnetic Resonance Microscopy. Группа авторов

manufactured by patterning thin copper tracks on a slender polymer foil so as to minimize the susceptibility-mismatch effects. The proposed LL was tested in a 9.4-T horizontal bore Bruker small animal MRI scanner. The scanner is equipped with a 72-mm-diameter volume coil. Three measurement scenarios were used to enable a comprehensive comparison of the performance. These scenarios are:

      1 MR imaging using the bare volume coil without any add-ons. This measurement served as the baseline to which the other scenarios can be compared.

      2 MR imaging with a broadband nonresonant LL integrated with the incubation platform and inserted in the volume coil.

      3 MR imaging of a resonant LL tuned, by a discrete capacitor, at the Larmor frequency and integrated into the incubation platform.

      Figure 1.5 A Helmholtz micro coil with a wire Lenz lens. [33] Nils Spengler et al. (2017), figure 03[p.008]/Public Library of Science (PLoS)/CC BY 4.

      Figure 1.6 Sensitivity enhancement of the micro Helmholtz coil due to Lenz lenses. The red curve indicates the reference signal-to-noise ratio (SNR) of the Helmholtz coil along the center line of the image when no Lenz lenses are used. The other curves show how the use of Lenz lenses boosts SNR of the sample region in the inner loop of the lens. The SNR enhancement ranges from 1.6- to 2.8-fold. [34] Nils Spengler et al. (2017), figure S1/Public Library of Science (PLoS)/CC BY 4.0.

In all experiments, a T1-weighted Flash sequence was applied to obtain an MR image of a mouse brain slice with a 0.5-mm thickness and an in-plane resolution of 100 × 100 µm from 16 averages over 8 min. The results, as demonstrated in Figure 1.7, show a significant enhancement in the imaging SNR due to LLs. More specifically, a broadband nonresonant LL achieved more than double the SNR of the volume coil, while an 8.5-fold SNR enhancement was obtained by the resonant LL.

      Figure 1.7 Top: Magnetic resonance (MR) compatible incubation platform for cultivating mouse brain slices. The platform was integrated with a broadband nonresonant Lenz lense (LL) (left), and a resonant LL (right). Bottom: MR images of a brain slice using the Bruker volume coil (middle), the broadband LL (left), and the resonant LL (right). [35] R. Kamberger et al. (2018), figure 06[p.13]/with permission from Elsevier.

      1.3 MR Microscopy and Neurotechnologies

“One cubic millimeter of cerebral cortex contains roughly 50,000 neurons, each of which establishes approximately 6,000 synapses with neighbouring cells” [36]. Thus any attempt to shed light on brain function using MRI or NMR will necessarily have to apply MR microscopy. A number of aspects, especially in brain research, are amenable to MR, especially those associated with possible brain interventions, and those involved in functional magnetic resonance imaging (fMRI) of the brain. We briefly consider these aspects next.

      1.3.1 Tissue Scaffolds and Implants

      Neurotechnologies rely on long-term implantable technical systems, in which technical materials come into direct contact with brain tissue, which mainly consists of neurons and a permeating vasculature. Two questions are of central concern:

      1 Do neurons permit intimate contact with the technical system?

      2 Do the materials of the technical system disturb subsequent MRI?

The realization that carbon, despite its hardness, is readily accepted by many cell types, also by stem cells, led to the exploration of microstructurable carbon as a tissue scaffold for studies in cell migration, network formation, and cell response [37]. In this study, an aqueous cryogel was steadily cooled, causing ice nucleation and crystal growth with morphological control, which upon thawing remained separated so that after drying a 3D polymer network remained. The network was pyrolized to yield an interconnected 3D carbon lattice, which in turn could be populated by neuronal stem cells, which allowed medium- to long-term studies of cell viability, confirmed by gradient echo MRI, see Figure 1.8.

Figure 1.8 Porosity and connectivity analysis of a neuronal scaffold using magnetic resonance imaging microscopy, comparing a cryogel with its pyrolized and hence shrunk carbon counterpart. [37] Erwin Fuhrer et al. (2017), figure 02[p.03]/with permission from John Wiley & Sons, Inc.

Moreover, the low susceptibility to warping of the magnetic field caused by the carbon scaffold at 11.74 T opened the door for the exploration of carbon as a neuronal implant [38]. In this study, the authors produced carbon brain implant microelectrodes on Kapton foil, by lithographically structuring a photopolymer followed by pyrolysis and embedding in durimine before release. The implants were then investigated for their MR properties, including force-induced vibrations using a specially constructed dynamic force sensitive probe head [39]. Compared with platinum electrodes, which sufficiently warped the MRI signal coming from neighboring voxels, carbon electrodes permitted the acquisition of unwrapped images right up to the microelectrodes, thus paving the way for studies in postoperation tissue recovery and wound healing.

      1.3.2 The Case of Epileptogenesis: Ex Situ Brain Slices and in Situ Histology

Brain-implanted microelectrodes target devastating diseases such as Parkinsons or epilepsy, and one of the primary questions in the field is whether it is possible to use noninvasive MRI to diagnose the disease, monitor recovery and temporal evolution of disease progress, and confirm correct system function of the implant. The hope is to translate diagnostic findings from mouse models to humans, despite certain differences, but with the benefit of using higher fields and hence higher-resolution techniques on model organisms than are generally available for humans. The gold standard for epileptogenesis is a kainate-initiated temporal lobe epilepsy mouse model established by Bouilleret et al. [40], for which a dedicated MR-compatible tissue slice environment was developed [35]. In this device, a freshly extracted brain slice 5 mm diameter, 500 µm thickness is maintained at physiological conditions (36°C, perfused, oxygenated) during MRI (Figure 1.7). Additional studies, which tracked the morphological changes of the brain slice over longer timeframes, were performed using a cryogenic vendor-supplied small animal coil [41], for the first time yielding confirmation of histological correlates with various MRI modalities. Using the same cryo-imaging coil, translation of these techniques to the full animal model yielded a comprehensive picture of the disease progression [42] and established a number of new measurement and postprocessing techniques, including high-resolution diffusion tensor imaging, see Figures 1.9 and 1.10.

Figure 1.9 Magnetic resonance (MR) and histological images of fixed hippocampal sections of two control animals (Section 1: A–D, with an adjacent section used for Golgi staining: E; Section 2: F–I). (A, B, F, G) Structural magnetic resonance imaging (MRI) depicts the main neuronal cell layers and tissue architecture. Comparison of diffusion tractography images (C, H) to corresponding NeuN and ZnT3 immunostaining (D, I) shows that regions containing parallel extending dendrites of principal neurons evoke corresponding diffusion-weighted imaging (DWI) streamlines. (E) Golgi staining of an adjacent section depicts the localization and orientation of principal cell dendrites (blue arrowheads in SR and ML) and parts of mossy fibers in CA3