06-01 Composition of MR Images
n the preceding chapters we have discussed the magnetic resonance phenomenon as such, the relaxation times, and the application of magnetic resonance to chemical analysis. However, the most important medical application of magnetic resonance is imaging (Figure 06-01).
The manner in which the spatial information is obtained in magnetic resonance imaging is referred to as the reconstruction technique. Images can be produced point-by-point, line-by-line, in slices, or in slices calculated from a whole volume (Figure 06-02).
Nearly all MR imaging techniques currently in use are either planar (slice) or volume techniques. In the former case, the MR experiment is restricted to a slice through the sample and is often referred to as a two-dimensional (2D) experiment since only two spatial dimensions have to be encoded.
Volume techniques spatially encode the whole volume; therefore, they are referred to as three-dimensional (3D) techniques.
The formation of an image involves the following procedures:
localization of the spins of interest;
excitation of selected spins;
spatial encoding of their signal; and
signal detection and reconstruction.
Each of these procedures, as well as their incorporation into a complete image formation experiment, will be discussed in detail over the following pages.
To create an image from a patient, the magnetic resonance signal from the nuclei has to contain information about where the nuclei are positioned in the patient. The MR equipment, as we have described it so far, does not provide us with any such information.
In MR spectroscopy experiments a sample is placed in a magnetic field which is shimmed to make it as uniform as possible. Now a particular molecule will give a signal of the same frequency at any point in the sample. Thus, any frequency changes observed in the Fourier-transformed signal reflect chemical shifts within the sample which can be used to create analytical spectra.