Spinal Cord
At present, the acquisition technique mainly employed in clinical routine for diffusion measurements is the so called single shot echo planar imaging technique. This technique is fast, the complete image is acquired in less then 0.1 s, and therefore robust against patient motion. However, for the application at the spinal cord, the echo planar technique suffers from strong susceptibility differences between the surrounding tissues.
The aim of this project is the development, implementation and optimization of appropriate imaging sequences for diffusion tensor imaging at the spinal cord. Several techniques may be employed in this regard. The “Parallel Imaging” and “Inner Volume” techniques can reduce the sensitivity to susceptibility differences by shortening the echo train length. Also, other acquisition techniques, as the so called “HASTE”-acquisition, offer new approaches. Since the diffusion weighting decreases the signal to noise ratio (SNR), all techniques must be optimized and validated in regard of the limited achievable SNR.
© dkfz.de
Normalized anisotropy of the diffusion tensor of a transversal image of the spinal cord. The anisotropy in white matter (axons) is larger than in gray matter (cell bodies). The typical butterfly-shape of the gray matter is visible.
Validation by Diffusion-Tensor-Phantoms
The signal to noise ratio (SNR) in diffusion weighted in vivo images is typically so small that image averaging is needed for a stable calculation of the diffusion tensor. In this regard, Monte-Carlo simulations are valuable for a better understanding of image noise on diffusion tensor imaging (DTI), but theoretical simulations cannot reflect the full range of interactions found in a true measurement. Also, sequence validation and comparisons between different MR scanners are not possible. On the other hand, true in vivo measurements are complex and the system is ill-defined.
Therefore, synthesised DTI phantoms that reflect the properties of neuronal tissue are of great value. To date, proposed synthesised phantoms could produce diffusion anisotropy, but had several restrictions. The diffusion properties of many phantoms where not in the range that is found in in vivo tissue. Other phantoms were difficult to produce and/or fragile. All were hand-made and hence hard to reproduce in a reliable fashion.
The aim of this project is the production of DTI-phantoms that show robust diffusion and MR properties matching the properties of white matter. By use of these phantoms, data can be acquired to investigates the influence of noise on the measurement of the diffusion tensor.
© dkfz.de
Left: Linear phantom in yellow shrinking tube. The water is between the polyamide fiber generates the measurement signal. The water diffusion is restricted by the fibers such that the diffusion becomes anisotropic.
Right: Measured colormap of the phantom. The direction of the principal eigenvector is color encoded and matches well to the fiber direction.