DWI benefits from the higher SNR, spatial resolution and accuracy of 3.0 T magnetic fields [26-30] (Fig. 7.4), despite the greater image distortion due to increased magnetic susceptibility effects compared with 1.5 T magnets (especially in areas of high susceptibility such as the paranasal sinuses). In such cases, parallel imaging or new techniques like PROPELLER (Periodically Rotated Overlapping ParallEL Lines with Enhanced Reconstruction) and radial fc-space filling can solve the problem of increased susceptibility (Fig. 7.5).
Diffusion studies adopt echoplanar techniques in which all phase-encoding steps are acquired with one RF excitation, thereby reducing the number of phase errors, which are responsible for susceptibility effects and image distortion. Parallel imaging reduces the number of phase errors by decreasing the number of phase encoding steps. Despite greater field inhomoge-neity, it enables control of geometric distortion and enhancement of spatial resolution up to 0.8 mm in plane.
Heightened SNR requirements are met in part by the higher sensitivity of 3.0 T magnets . Using parallel imaging at 3.0 T, DTI data sufficient for brain fibre tracking in healthy subjects can be obtained in < 2 min with 2 mm slice thickness, 700 s/mm2 b factor, six motion probing gradient directions, and no averaging (number of averages = 1) .
A diffusion sequence based on single-shot FSE has recently been proposed to address the problem of susceptibility, as this sequence is not affected by the spatial distortions which usually limit conventional acquisitions with EPI .
DTI with PROPELLER shows a significant potential to reduce the number of weighted acquisitions, avoid ambiguity in reconstructing diffusion tensor parameters, increase SNR, and decrease the influence of signal distortion .
The linear increase of the MR signals with increased B0, entailing a reduction of partial volume effects, is particularly beneficial for DTI-based fibre tracking. In fact, the reconstructed DTI trajectories reflect more accurately the underlying axonal fibres, particularly crossing or bifurcating bundles .
High-field strength MRI is very sensitive to hyperacute ischaemic lesions also at lower b values (500 instead of 1,000 as in 1.5 T systems) (Fig. 7.6) , where as tumours must be investigated at high b values to distinguish the (hyperintense) lesion from the surrounding (hypointense) vasogenic oedema (Fig. 7.7).
The higher SNR makes it possible to increase the b value to enhance the anisotropic effect, thus improving the quality of more complex evaluations like anisotro-py, DTI and TI, which are poorly displayed on 1.5 T images, even with multiple excitations, due to their insufficient SNR  (Figs. 7.8-7.11). Using a 3.0 T system, these diffusion techniques seem to provide more informative data for the study of brain tumours, especially to assess the involvement of peritumour white matter,
which may exhibit oedema and/or neoplastic infiltration. In particular, subtle white matter disruption can be identified using DTI in patients with high-grade gli-omas. Such disruption is not seen in association with metastases or low-grade gliomas, despite the fact that these tumours produce significant mass effect and oedema. In the latter case, white matter fibres are merely compressed and dislocated, not infiltrated or disrupted as in high-grade gliomas. The changes detected on DTI may therefore be due to tumour infiltration. For this reason, DTI can be a valuable method for detecting occult white matter invasion by glioma, thus providing useful information for planning treatment .
Fig. 7.7 c b value = 3,000. At the higher b value (c) the tumoral core (hyperintense) can be distinguished from the peripheral vasogenic oedema (hypointense)
■BBnQQaass==23n33=33S2aanDBBBBBBBaanc iMHggn5BBB3saB3BBaaaaagE3nnMWHanc ^BsaaaaaaBasaajaapnBa—i mil il il il I I I I h mil Baaaa^Qicana i aaamr^aanaamm maanz&znaammm
Was this article helpful?