Application of MRI in small animal models allows for a more closer look on basic physiological processes and the high sensitivity at our 16.4T animal scanner refines even more the precision of imaging and spectroscopy relative to human studies at lower field strengths. My job is the technical realization and optimization of functional imaging protocols on anesthetized rodents and the active support of other projects at the animal scanner (e.g. contrast agent development, RF-coil engineering, localized in vivo spectroscopy, 17O-imaging). The running projects I am most involved in are:

• High resolution fMRI in the rat at 16.4T
• In vivo MR-microscopy in the mouse abdomen at 16.4T: detection of isolated microscopic susceptibility effects during fast movement with big amplitude

1. High resolution fMRI in the rat at 16.4T

Introduction

Spatial specificity of the blood oxygen level dependent (BOLD) contrast is expected to increase with the magnetic field strength, leading to better localization of the observed vascular effect to its neural origins [1]. To benefit from these changes [2], physiological noise has to be minimized. This is usually achieved in studies on small animals by using controlled anesthesia and by pro- and retrospective motion correction. Particularly at high-fields, the increased sensitivity can be invested in improved spatial resolution, thereby reducing the ratio of physiological to thermal noise. Additionally, novel anesthesia protocols for animal preparation and imaging, which account for pain-related stress effects on the physiological state of the brain [3], can improve reproducibility of fMRI-studies in rats.

Goals

The goal of this project is to generate BOLD-maps with very high spatial resolution in anesthetized rats at 16.4T. Stress for the animals should be minimized using an optimized surgery and anesthesia protocol.

Methods

BOLD mapping with echo planar imaging (EPI) was performed on the rat forepaw stimulation model with intensive physiological monitoring. To guarantee unbiased results, a minimally invasive artery catheterization procedure was developed and the stress level with different anesthetics was analyzed by quantifying plasma corticosterone.

Initial results

Results from gradient-echo (GE) and spin-echo (SE) EPI datasets with overlaid BOLD activation maps are presented in Fig.1. Isotropic resolutions of 150 μm or inplane resolutions down to 50 μm can be obtained. Improved localization reduces partial volume effects and activation-related changes up to 40% are detected in single voxels. Stress during preparation is shown to be significantly reduced for anesthesia with propofol and remifentanil, especially when compared to the mostly used isoflurane.

Initial conclusion

Enhanced SNR at 16.4T was invested in spatial resolution, resulting in improved specificity of the BOLD-effect. A novel anesthesia protocol for the surgical preparation of rats preceding fMRI was developed, which can further minimize confounding physiological effects.

References

1. Seehafer JU, Kalthoff D, Farr TD, Wiedermann D, Hoehn M (2010) No increase of the blood oxygenation level-dependent functional magnetic resonance imaging signal with higher field strength: Implications for brain activation studies, The Journal of Neuroscience 30:5234-5241.

2. Pflugfelder D, Vahedipour K, Uludag K, Shah NJ, Stöcker T (2011) On the numerically predicted spatial BOLD fMRI specificity for spin echo sequence, Magnetic Resonance Imaging 29:1195-1204.

3. Ferris CF, Stolberg T (2010) Imaging the immediate non-genomic effects of stress hormone on brain activity, Psychoneuroendocrinology 35:5-14.

Figure 1: High-resolution BOLD activation maps in the rat brain at 16.4T. MR-method, resolution and the repetition time per volume are indicated.

Figure 2: Corticosterone concentrations during artery catherization at different blood sampling time-points for CD, Wistar and Fischer-rats in separate histograms as measure of the stress level.

2. In vivo MR-microscopy in the mouse abdomen at 16.4T: detection of isolated microscopic susceptibility effects during fast movement with big amplitude

Introduction

Magnetic resonance microscopy (MRµ) is used for visualizing sample structures around and below (100µm)³. MRµ of fast moving samples is challenging and, hence, in vivo studies on animal models are restricted to the head or extremities that can be immobilized. The abdomen of an anesthetized mouse moves more than 100 times per minute with an amplitude of several millimeters, because of respiration. Thus, in case of MRµ in the mouse abdomen, the target completes in about 500ms a trajectory, that is more than an order of magnitude longer than its own dimensions. In case of this study, the targets were the sites of insulin production, islets of Langerhans in the pancreas. These microscopic structures possess no intrinsic structural contrast to differentiate them from the surrounding pancreatic tissue. For detection with MRµ, they have to be labeled with islet-specific MR-contrast agents.

Goals

The aim of this study was the in vivo visualization of single native pancreatic islets (R~50µm) marked with intelligent contrast agents.

Materials and Methods

The required signal per unit time was achieved by performing experiments at 16.4T with custom-built coils, dedicated for mouse imaging. T2*-contrast was generated by the contrast agent, which was composed of biospecific cargo molecules attached to ferromagnetic carbon coated Cobalt nanoparticles. To avoid effects of motion, two different acquisition strategies were used: 1. Respiration triggered signal sampling, which resulted in the same position and size of the inner organs during the read-out of all k-space lines in an imaging slice; 2. Massive averaging of full k-space acquisitions, which resulted in the statistical mean k-space over all interim states in the respiratory cycle. Because of short repetition times, the second strategy did not allow full abdomen coverage.

Results and Discussion

The triggered experiment provided an image resolution of 100x100x300µm³ in 9:20min, and susceptibility weighted image reconstruction had to be performed to resolve the isolated microscopic susceptibility effects (Fig.1). The continuous motion corrected acquisition yielded 66x66x100µm³ image resolution in 11min and local contrast from pancreatic islets was detected directly (Fig.2). The later strategy was tested in vivo at 7T, but could not reproduce the same spatial and temporal resolution as on 16.4T, and the visualization of pancreatic islets was not possible.

Conclusion

Sufficient signal and contrast efficiency, provided by the 16.4T MR-system, specialized coils and a ferromagnetic contrast agent, were essential for successful MRµ in the mouse abdomen in vivo. For sharp contrast and direct islet-visualization, a motion corrected 2D acquisition strategy was used. For coverage of the whole pancreas and detection of relative changes in islet-contrast, respiration triggered sampling combined with susceptibility weighted image reconstruction can be a feasible option.

Figure 1: Triggered 2D-FLASH images of a mouse abdomen acquired with a surface coil. TR~600ms (respiration period), TE=1.9ms, 100x100x300µm³, NA=2. (a) Magnitude image. (b) SWI. Pancreatic tissue is marked (red arrows); contrast enhanced regions are highlighted and two zoom-ins are presented as insets.

Figure 2: Motion corrected FLASH images of the mouse abdomen (volume coil). TR=15ms, TE=2.75ms, NAE=250, 66x66x100µm³, slice gap 2mm, FOV=10x23mm, horizontal readout, ACQ=11min. Green = liver; yellow = spleen; red = pancreas; violet = kidney.