T2Weighted Perfusion

SE-based DSC-MRI exhibits some distinct physical and physiological differences as compared with the GE method. Simulation studies as well as in vivo experiments have demonstrated that SE-based measurements are less sensitive to microvasculature than GE-based techniques. In computer simulations, the sensitivity of a GE-based method is relatively independent of the size and distribution of vessels and capillaries within a given voxel, whereas SE-based methods show a maximum sensitivity for vessels between 5 and 10 im; the size of capillaries in the human brain (Boxerman et al. 1995). Although the selective sensitivity to microvascularization of SE methods is advantageous, DSC-MRI is often performed by the GE method due to better coverage of the brain (given the same repetition time, it is possible to obtain more slices during a single acquisition). Furthermore, due to the susceptibility contrast sensitivity of GE sequences for all dimensions, the contrast dose used in the GE experiment is typically only half of that used for SE (Simonsen et al. 2000) On the other hand, since the SE sequence has the inherent advantage of being less sensitive to differences in magnetic susceptibility at air-tissue interfaces, the method offers better anatomic precision and detail in clinical studies. Parametric analysis using

SE-based EPI techniques are predominantly based on the susceptibility effects rather than changes in the T2 relaxation rate. This is in fact a consequence of the relatively long sampling period of the EPI acquisition technique. The susceptibility contrast arising from compartmentalization of the contrast agent is therefore used to determine relative tissue and relative arterial concentration levels according to Eq. 3, allowing subsequent calculations of rCBV, rCBF, and MTT in the same manner as described in Chap. 4.2.1 (Fig. 7.1).

Importantly, several studies have indicated that the SE- and GE- (both EPI) based methods are comparable with regard to the relevant parameter for in vivo studies (Speck et al. 2000; Weisskoff et al. 1994). In fact, the SE method shows smaller standard deviations than the GE method for small regions that are adjacent to or contain large vessels (Speck et al. 2000). However, great care must be taken when comparing parametric information based on SE DSC-MRI with those obtained by GE DSC-MRI in tumors. It has for example been shown that the tumor rCBV obtained

Gradient Echo Spin Echo Anatomy rC BF Map rCBF Map survey

Gradient Echo Spin Echo Anatomy rC BF Map rCBF Map survey

Fig. 7.1. rCBF maps calculated from the gradient-echo, the spin-echo, and the corresponding anatomic inversion recovery images. It was found that the appearance of large vessels was markedly reduced in the spin-echo images, resulting primarily in a representation of capillary perfusion. [Reprinted with permission from SpEck et al. (2000)]

Fig. 7.1. rCBF maps calculated from the gradient-echo, the spin-echo, and the corresponding anatomic inversion recovery images. It was found that the appearance of large vessels was markedly reduced in the spin-echo images, resulting primarily in a representation of capillary perfusion. [Reprinted with permission from SpEck et al. (2000)]

with the GE method was significantly higher than the SE method in the high-grade gliomas, but not in the low-grade gliomas. Simulations of the microscopic susceptibility variation in SE and GE imaging have subsequently demonstrated that the SE method produces a relaxivity peak for capillary or smaller sized vessels, and that SE-based DSC-MRI demonstrated a capillary blood volume that is more physiologically interesting than total blood volume obtained from conventional GE-based DSC-MRI (Weisskoff et al. 1994).

Another method to integrate the SE technique into DSC-MRI of the brain is the half-Fourier acquisition single-shot SE (HASTE), enabling heavy T2-weight-ing during its imaging interval of approximately 12 s (Miyazaki et al. 1996). The HASTE method has successfully demonstrated its availability to measure cerebral microvascular hemodynamics (Koshimoto et al. 1999). In addition, because of HASTE's low susceptibility to artifact, image degradation was effectively eliminated, but the major drawback of the HASTE method is its limitation to obtain more than one or a few slices as compared with the whole brain coverage of the EPI methods.

In general, SE-based DSC-MRI is promising with respect to the detection of the cerebral intravascular effects. Unfortunately, SE methods are only sensitive for T2* changes in and around capillaries, and not around larger size venules and arteries (Boxerman et al. 1995). As a consequence, signal changes during passage of the bolus are lower, and SE techniques are therefore hardly used in the clinic for bolus tracking. Nevertheless, this may change with the use of very high field magnet strengths.

0 0

Post a comment