Diffusion measurements using MRI
Diffusion imaging was first conducted in vivo in the mid-late 1980:s and a few years later it was discovered that this technique allowed localization of ischaemic stroke in the early stage (Figure 1.) (1), something that until that point of time not had been possible using conventional MRI or computerized tomography (CT).
Figure 1. A diffusion weighted image and a diffusion coefficient map showing an acute ischaemic stroke lesion as an area of bright signal intensity in the DW image and an area of low signal intensity in the diffusion coefficient map.
Diffusion measurements are based on a labeling procedure where water molecules are marked with respect to their initial position through the application of a magnetic field gradient pulse of high amplitude. After a short period of time, during which the water molecules are given time to relocate, a second gradient pulse is applied. The resulting MR signal will now depend upon the combined effect of the two gradient pulses (the diffusion sensitivity), the diffusion speed and the time during which the molecule were allowed to move around. At the end of the experiment, the stationary molecules will have a strong MR signal, whereas diffusing molecules will show a reduced signal. If two experiments are carried out, with different amount of diffusion sensitivity, it is possible to determine the diffusion rate, expressed in terms of the diffusion constant. For in vivo imaging several factors, such as restricted or hindered diffusion as well as perfusion effects, will contribute to the observed diffusion rate and the term apparent diffusion coefficient (ADC) was introduced to account for this. Since the diffusion speed only can be determined in one direction at a time, a clinical diffusion MRI often includes measurements in three orthogonal directions. Examples are given in Figure 2.
Figure 2. DW images obtained with sensitization made through plane (CC), left-right (LR) and anterior-posterior (AP), clearly demonstrating the anisotropy effects caused by nerve fibre orientation.
Once the images displaying the fractional anisotropy and the preferred diffusion direction has been created, it is possible to perform so called fiber tracking (Figure 4), or diffusion tensor tractography (DTT). In order to perform three dimensional fiber tracking it is preferable to acquire isotropic voxels (e.g. 2×2×2 mm) and the final image quality will depend upon the number of diffusion directions sampled, the result will improve with increasing number of directions.
Figure 4. A set of images displaying the principal diffusion encoding direction in an RGB color scale (the rightmost image), a corresponding image where the content of all pixels has been replaced with rectangular boxes, and below a blow up of the lower part of the corpus callosum. To the right three different views of the fiber bundles in the corpus callosum.
No comments:
Post a Comment