In MRI, tomographic images of tissue slices having defined orientation, thickness and location are obtained. The slice selection process is accomplished by using a frequency-selective RF pulse during the application of a magnetic field gradient. A frequency-selective RF pulse permits the highly selective excitation of MRI signals having a narrow band of Larmor frequencies. The presence of a magnetic field gradient ensures that this narrow band of Larmor frequencies will correspond with a narrow tissue section. Protons that would generate signal outside of the narrow band of Larmor frequencies (i.e., those located beyond the edges of the narrow tissue section) are unaffected and do not produce signal. Figure 5 illustrates that an axial tissue section may be selected for subsequent imaging by using a frequency-selective RF pulse in the presence of a magnetic field gradient oriented along the inferior
to superior axis. Both the slice thickness and the slice location may be freely selected without physically moving the human subject or physically moving components of the scanning equipment relative to the subject. This is done by making suitable adjustments in the characteristics of magnetic field gradient or the frequency-selective RF pulse. Accordingly, the moving parts of MRI scanners tend to be limited to that needed for moving the subject into the appropriate part of the magnetic field at the beginning of the examination. Subsequently, neither the subject nor the scanning equipment moves during the image acquisition process. Furthermore, because a field gradient can be created along any arbitrary anatomic axis, it is possible to select slices in any of the three principal anatomic sectional planes (sagittal, axial, and coronal) or in any arbitrary oblique imaging plane without physically repositioning the subject or physically reorienting the imaging equipment (Fig. 6).
The obvious need to visualize anatomy in three dimensions is usually met through the use of mutlislice imaging. Figure 7 displays a series of 15 axial images that were obtained from tissue sections 3 mm thick. The conventional practice is to produce collages showing individual two-dimensional sectional images as a means of visualizing three-dimensional (3D) anatomy. If the sections can be made sufficiently thin, it is possible to create a volume rendering that displays the cortical surface anatomy. Conventional multislice MRI for human subjects is limited to a slice thickness of about 3 mm. This is relatively thick for volume rendering purposes; therefore, when volume rendering is planned, special 3D acquisition techniques are employed in which more than 100 slices having slice thickness of about 1 mm are used. Figure 8 provides an example of volume rendered images that display brain surface anatomy from a variety of perspectives. In addition to viewing surface anatomy in the manner shown, it is also possible to section the digital representation of the brain volume and produce two-dimensional images in any chosen sectional plane. This is done retrospectively using purposely designed software that sections and displays images under user control. The volumetric nature of the image data also readily lends itself to studies in which it is desired to measure the volumes of specific neuroanatomical structures that can be visualized in the images. As a result, 3D acquisition techniques are often used in research studies in which quantitative evaluation of neuroanatomic volumes are sought. The volumetric
Figure 8 Volume rendering of three-dimensional MRI.
nature of the image data is also appropriate for the planning of neurosurgical procedures such as stereotactic biopsy. Accordingly, 3D volume MRI acquisition is often performed as part of the preoperative evaluation of patients for whom neurosurgical procedures are contemplated.
Was this article helpful?