Neuroimaging of neck pathology.

Computed tomography (CT) scanning in the head and neck region has become an integral part of the preoperative evaluation of tumor in defining the full extent of anatomic involvement, and the relation of the tumor to vital vascular and neurologic structures. CT utilizes the same ionizing X rays as plain-film radiography, except the X-ray beam is a thin, collimated linear beam, single but more often multiple; in plain films, a single beam is used, which disperses outward. In CT, the beam attenuated by the object being studied then is detected by a crystal detector and converted to electric energy. In plain film, the attenuated beam then strikes a photoemulsion silver-based film, causing a photochemical reaction, which can then be processed into a photograph. Thus, the CT scan is not a real image of a real object, whereas the plain-film radiograph is. Basically, the object being studied casts a shadow (absorbs the X ray so that it does not penetrate to the film) of a real object. In CT, the final picture is a reconstructed computer simulation. Nevertheless, the simulation does accurately reflect the real anatomy. Unlike plain film, CT scanning takes thousands of small images and actual X-ray beam exposures compared with the one shot in plain film. Because the beam is so thin, however, the overall tissue absorption is about equal to, and sometimes less than, the amount of radiation exposure encountered in plain-film radiography. To generate a picture utilizing a matrix, there must be a way to spatially encode the thin X-ray beams of CT. This is accomplished in part by the rotating ring, with single, multiple, and radial beams, by incrementing in 1� intervals the X-ray tube and the crystal detectors around the object being studied, and taking pictures at each incremental rotation. In MR imaging, the spatial encoding is accomplished by applying small, linear, magnetic gradients, and thus frequency encoding, in the X direction, and phase encoding in the Y direction. The current CT technology then utilizes continual scanning but incrementing in a helical or spiral fashion to reduce further the CT scanning times, as opposed to the earlier scan utilizing one slice, stop, move table, and repeat. The latest technology places the crystal detectors in arrays or banks, rather than the linear-or circular-oriented single row of detectors. None of the new technologies increases anatomic resolution but instead reduces scanning times, which benefits CT angiography or cinefluoroscopic CT. The electric signal is then analyzed by Fourier transformation, which further adds information about spatial encoding. The spatially encoded electric signal, by means of back projection, can then be used to determine how bright or dark each pixel of the picture matrix should be. Resolution as high as a 1024 · 1024 matrix can then be created for temporal bone imaging. More often, a 512 · 512 matrix is used for most CT scanning of the head and neck. To generate a gray-scale image, a zero point needs to be established. CT intensity readings are assigned Hounsfield units (HUs), with the zero value representing water and cerebrospinal