Biomedical Engineering Reference
In-Depth Information
so, they emit rf radiation at the same Larmor frequency. I think of it
as the spins being excited by the initiating rf field and then giving out
little rf “yelps” as they de-excite. The sound of the yelp, that is its
frequency, depends on the magnetic field strength; it is the Larmor
frequency. If a second receiver rf coil is placed nearby, it can detect
the emitted rf radiation and thus “sense” the presence of protons.
If the tissue sample (e.g., a patient) were exposed to a uniform
magnetic field, then all the protons within the sample would precess
at the same frequency, they would all be excited by the same rf signal,
and would all contribute to the received rf signal. That is, there
would be no information about the spatial location of the protons.
Imagine, on the other hand, that an inhomogeneous magnetic field
could be applied in such a way that each small volume of tissue
(voxel) experienced a different magnetic field from all others. Then,
the Larmor frequency of the protons in a given voxel would be
different from that of all other voxels since the magnetic field would
be different from that in all other voxels. If an rf field were applied of
exactly that voxel's particular Larmor frequency, then only the
protons in that voxel would be excited and then the detected rf signal
would come from the de-excitation of those protons alone. By
sequentially varying the applied rf field in time, one could obtain a set
of detected signals, each unique to a single voxel. That is, one would
have complete 3D information as to the behavior of the protons in
each voxel, independently.
Unfortunately, the laws of physics (in particular, Maxwell's laws of
electromagnetism) do not allow one to design a magnetic field whose
strength is different at each point in 3D space. But, one can easily
achieve partial spatial information by applying a magnetic field
gradient across the tissue sample, thus creating planes of different
magnetic strength. The received rf signal in that case would come
from all the protons in a plane. This is, in fact, the first step in
obtaining 3D information in MRI. How the remaining two dimen-
sions of decoding are disentangled is beyond the scope of this
short summary. Suffice it to say that, by applying various magnetic
field gradients in sequence, and by processing the received signals,
one can analyze the signals so as to isolate the rf response of the
protons in each voxel of a 3D array of voxels. Since the information
is truly three dimensional, it is as easy to display a sagittal or coronal
section as it is to display a transverse section.
