MRBased Techniques

MRI imaging is the most active research area for new techniques. Some of these techniques, especially diffusion imaging, are already being used by neurosurgeons to discriminate tumor types. Others, such as magnetic resonance spectroscopy (MRS) and functional MRI (BOLD), were thought to be promising in tumor characterization in the past, but have not been shown to be really effective in clinical studies. This section focuses on three techniques: diffusion imaging, MRS, and dynamic color mapping. These are particularly attractive because they do not require new hardware and can be performed on current clinical MR scanners. In contrast, other interesting "new" techniques discussed in this chapter, such as PET and SPECT, typically require large investments in scanners and nuclide instrumentation.

Diffusion-Weighted MR Imaging

Diffusion-weighted MRI, which is similar to regular MRI and employs the same scanner, measures the potential of water molecules to move and is based on the random, microscopic movement of water molecules in tissue. As water molecules collide, they spread out (diffuse). Normally (e.g., in normal brain gray matter), this motion is unrestricted, which leads to spin dephasing and attenuation of the MR signal. However, if the motion is restricted, which can happen if the tissues consist of parallel fibers, such as nerves or skeletal muscles, or if there is a restricted extracellular space, such as in high-cellularity tumors, the MR signal will be less attenuated. Diffusion-weighted images are MR images in which the signal intensity shows the MR attenuation caused by diffusion. On these images high diffusion is shown as black and low diffusion (the optic nerve and extraocular muscles) as white. The resulting MR signal can be normalized in several ways, leading to images of differ ent "diffusion signals," known by acronyms such as TDC (true diffusion coefficient), ADC (apparent diffusion coefficient) or Dav (average diffusion constant), which are not discussed here.

Diffusion imaging measures function on a microscopic scale. Diffusion in tissue is largely determined by microscopic features such as density and type of cells and intracellular matrix, since these structures impede the free movement of water molecules to a greater or lesser extent; compare this with the common forms of MRI such as T1, T2, and proton imaging, which measure processes on the atomic scale. Theoretically, changes in diffusion may more directly reflect changes occurring within and between cells.

In brain gliomas, diffusion-weighted imaging was found to differentiate between high and low cellular-ity as found on histopathological examination (different ADCs) in a prospective study.21 Cellularity is an important histologic determinant of glioma malignancy. Histopathological cellularity of lymphomas was found to be predictable by means of diffusion im-aging.22 In a prospective study of brain meningiomas, diffusion-weighted imaging was found to differentiate between aspecific and malignant meningiomas on the one hand and benign meningiomas on the other, with malignant and aspecific meningiomas showing low Dav on imaging.23 In a prospective study of soft tissue tumors, diffusion imaging could predict the malignancy of soft tissue tumors as determined by histo-pathology, with benign tumors (such as leiomyoma and schwannomas) showing high TDC and malignant tumors (such as liposarcoma and leiomyosarcoma) showing low TDC.24

Magnetic Resonance Spectroscopy

Magnetic resonance spectroscopy uses the same concepts as MRI, but instead of imaging the spatial relationships of tissues, the concentrations of specific chemical compounds within the tissue are imaged. The radiofrequency pulse sequence of the scanner produces a signal decay, the amount of which is ultimately determined by the chemical relationships in the tissue. These series of signals can be transformed into a spectrum, where the concentration of different chemical compounds is imaged via different color intensities. By using MRS, it is possible to image the concentration of different compounds in a particular region of interest such as the brain or the orbit. The most interesting compounds for tumor characterization are choline, creatine, N-acetyl aspartate, and N-acetylaspartyl glutamate.25 Choline and creatine are present in most cells, whereas N-acetyl aspartate and N-acetylaspartyl glutamate are localized predominantly in neurons. The ratios between these compounds help to indicate the type of tissue in the tumor. The proliferative capacity of gliomas as measured

FIGURE 11.1. Anatomy of the right orbit showing the results of new research. The orbit has been opened, and the globe and Tenon's capsule have been removed. In the superior part of the orbit, the extraconal fat and the periorbital have been removed. The pulleys (faintly bluish fibers) insert on the orbital walls via the periorbital (blue arrow) but are stable because of their interconnections (black arrow) even when the pulley insertions are removed from the wall. The intraconal fat (yellow arrow) has fluidlike behavior, even though it is part of the same connective tissue network as the pulleys. (With permission from M. D. Abramoff, "Objective Measurement of Motion in the Orbit" [Ph.D. thesis]. Transatlantic Publishing, New York, 2001.)

FIGURE 11.1. Anatomy of the right orbit showing the results of new research. The orbit has been opened, and the globe and Tenon's capsule have been removed. In the superior part of the orbit, the extraconal fat and the periorbital have been removed. The pulleys (faintly bluish fibers) insert on the orbital walls via the periorbital (blue arrow) but are stable because of their interconnections (black arrow) even when the pulley insertions are removed from the wall. The intraconal fat (yellow arrow) has fluidlike behavior, even though it is part of the same connective tissue network as the pulleys. (With permission from M. D. Abramoff, "Objective Measurement of Motion in the Orbit" [Ph.D. thesis]. Transatlantic Publishing, New York, 2001.)

by histopathology and response to treatment was found to correspond to the choline/creatine level in a prospective study.26 Medium to high ratios of choline to creatine have been used as a marker for the presence of actively proliferating tumor cells, whereas decreases in the overall levels of choline, creatine, and N-acetyl aspartate, as well as increases in lipid/lactate proton resonances, were found to correspond to ne-crotic processes.27

MR Dynamic Color Mapping

MR dynamic color mapping uses a specific property of the orbit.28 All tissues in the orbit are involved in motion during gaze changes, and changes in tissues change the motion of these tissues (Figure 11.1). For example, a tumor originating in a rectus muscle will move along the same trajectory as that muscle, even when it is spreading through other tissues. Once it has grown invasively into another tissue, however, the motion of the tumor starts to mimic the motion of the invaded tissue. If a tissue is adjacent to another tissue, it may or may not have similar motion, but the more connected the tissues become, the more their motion becomes similar.

Dynamic color mapping involves two stages: a stage of cinematic MR, where sequences of MR images are acquired with the eyes moving along different gaze positions, and a second stage, during which the motion fields in these image sequences are computed and imaged. Though MRI is able to image realtime motion (called dynamic MRI), it is at present too slow to do so in the comparatively small orbit at acceptable resolution.15 Therefore, motion imaging uses cine acquisition. Here, the tissues are allowed to move over the full range of motion with small increments. After every increment (stop), a new image is acquired (shoot). The result is a sequence of images (Figure 11.2). The difference between dynamic and cine acquisition is relative, not absolute, because in the limit where the time increments are very small, cine acquisition is the same as dynamic acquisition. The term cine derives from "cinematic": the stop-shoot technique is comparable to cartoon animation.

Cine acquisition has been used to image the role of connective tissue in determining extraocular muscle paths,29-31 the role of connective tissue ligaments in eyelid motion,32 and the optic nerve path.33 The motion in the anophthalmic socket has also been studied.34 Cine MR requires gaze sequencing to allow the patient to gaze over the specified range. An example is the Snow White machine (Figure 11.3), consisting of a transparent acrylic half-pipe fitted snugly in the scanner bore. On the inside is a row with nine fixa-

FIGURE 11.2. Typical frame of an orbital cine MRI sequence. The corresponding position of the eyes is shown. (With permission from M. D. Abramoff, "Objective Measurement of Motion in the Orbit" [Ph.D. thesis] Transatlantic Publishing, New York, 2001.)

FIGURE 11.3. Snow White machine fixation aid. (With permission from M. D. Abramoff, "Objective Measurement of Motion in the Orbit" [Ph.D. thesis] Transatlantic Publishing, New York, 2001.)

tion marks indicated with numbers 1 through 9. This sequence is horizontal, and the marks are 8° apart if the rotational centers of the eyes are 200 mm from the device. There are also two rows, vertical with respect to the patient's face, one 20° to the left and one at 20° to the right of the straight-ahead fixation mark. The Snow White machine was left transparent to minimize reactions of claustrophobia (Figure 11.4). To compute and visualize the two- and three-dimensional motion of the orbital tissues, there are specific computer algorithms.35,36 The resulting images use colors to specify the direction of the moving tissues and also the magnitude of the motion (Figure 11.5). These images have been validated in recent studies.35-37

The use of two- and three-dimensional MR dynamic color mapping revealed that the orbit behaves similarly to the organ of gaze and that the so-called soft tissue in the orbit is actually quite rigid in the anterior, forming a skeleton, while the soft tissue in the posterior part of the orbit is almost fluidlike.38,39 The cause of ocular motility changes after some forms of decompression surgery were also found.

MR dynamic color mapping has been used in the orbit to discover microscopic neuromas, small and painful nerve tumors in the optic nerve stump. Neuromas, which can arise after enucleation and implantation in the orbit because of constricted motion, can cause persistent pain that is not responsive to other treatment (Figures 11.6 and 11.7).40 These tumors were suspected because the MR dynamic color mapping showed restricted motion of the optic nerve stump and attachment of the stump to the scleral cover of the implant.

MR dynamic color mapping may also help in determining the origin of orbital tumors, as was shown in two patients in whom the tumor was found to be originating in the rectus muscle because it moved in the same direction and at the same magnitude as that muscle and not with the optic nerve to which it was adjacent.37

In summary, a considerable amount of evidence exists for diffusion-weighted imaging, and a somewhat lesser amount for magnetic resonance spectroscopy and MR dynamic color mapping, to suggest that these may be promising techniques to characterize orbital tumors and help decide the which-tumor-is-it question.

0 0

Post a comment