2d Tof Mrv Scan Parameter

Fig. 35. Example of a CSF flow study using the phase contrast technique to determine whether there is increased flow through the aqueduct. rf, radiofrequency.

velocity and is therefore a direct measurement of flow, compared with TOF, is that three dataset acquisitions are necessary in three orthogonal directions, therefore tripling the imaging time (30). The advantage of this technique is that it is a true measure of velocity and is invaluable in distinguishing slow flow from thrombus (in which short T1 thrombus and slow flow will appear as high signal on TOF) and that it is also the method used for CSF flow study primarily for determining shunt placement for patients suspected of normal pressure hydrocephalus (31) (Fig. 35).

Even with the various advantages of the phase-contrast technique, TOF remains the method of choice for most MR vascular flow studies (not CSF flow) primarily because of the time-saving factor. As one can expect, this method does not distinguish the direction of flow and therefore both arterial and venous flow will be depicted on the images. To obtain only arterial (MRA) or venous (MRV) flow, a saturation band is used that is essentially a thin slab of gradient that is turned on and placed proximal to the direction of the flow that one wants to suppress. The flowing spins are therefore saturated before entering the imaging volume and will not contribute to any signal formation. For MRV, the saturation band will be placed proximal to the common carotid arteries; for MRA, it is deployed at the cranial end of the head to suppress venous spins that flow in the caudal direction.

Both 2D and 3D TOF techniques can be employed for MRA; each has its strength and weaknesses. Two-dimensional TOF (Fig. 36) is essentially made up of sequential thin slices oriented perpendicular to the flow of the vessel. Because of the thickness of the slices, 2D TOF is sensitive to slow flow (applicable for evaluation of carotid stenosis) but requires maximum gradient strength and thus pays the price of longer TE and greater susceptibility to spin dephasing. In comparison, 3D TOF consists of a relatively thick slab that allows

Fig. 36. A typical 2D TOF pulse sequence.
Intra Voxel Dephasing Phenomenon
Fig. 37. 2D (left) and 3D (right) TOF MRA of the neck showing less intravoxel dephasing or signal loss at the carotid bulb in the latter, avoiding potential erroneous diagnosis of carotid artery stenosis.

for better SNR, a shorter TE, and less intravoxel dephasing (Fig. 37). On the other hand, the drawback is the risk of insensitivity to slow flow conditions: by the time the spin has traveled from one end of the slab to the other it has already become saturated. Clinically, 3D TOF of the circle of Willis is used mostly for the diagnosis and screening of aneurysms and vascular malformations; 2D TOF MRA or MRV is employed for evaluation of carotid bifurcation stenosis and suspected cases of venous sinus thrombosis, respectively (Figs. 38-40).

Both TOF methods are also inherently insensitive to the problem of in-plane flow, i.e., flow that is parallel to the imaging plane. To counteract this, the

Basil Artery Bifurcation Mri
Fig. 38. 3D TOF MRA of the circle of Willis depicting right middle cerebral artery bifurcation and basilar tip aneurysms in two different patients (left and right).
Circulation And Tof
Fig. 39. Source and reconstructed 2D TOF images of a normal MR venogram in three orthogonal planes (clockwise from top left).

administration of gadolinium DTPA a contrast agent that shortens T1, is used to strengthen the signal from the longitudinal relaxation of blood as it remains within the circulation. This also permits the acquisition of data in the coronal or sagittal or other oblique planes, allowing for more coverage with fewer slices and less time (for example, from the aortic arch all the way to the intracranial vessels with the source images acquired in the coronal plane) and eliminating the issue of in-plane flow (Fig. 41). One of the occasional disadvantages of contrast MRA is that the indiscriminant T1 shortening results in a technique that

Mrv Thrombus
Fig. 40. Axial CT (high attenuation), T1 noncontrast MR (high signal), and 2D TOF MRV images of a right lateral sinus thrombosis (clockwise from top left).

emphasizes the opacification of the vessel and in essence (like computed tomography angiogram) may decrease the functional aspect of MRA. In other words, if there is high-grade stenosis at the carotid bifurcation, there is often decreased flow signal distally, especially for the intracranial vessels in the non-contrast setting. With the T1 shortening effect of gadolinium, these intracranial vessels may show signal, thereby underestimating the severity of disease. In addition, as contrast flows into the venous system, the timing of the bolus administration and data acquisition becomes crucial in the success or failure of the study (Fig. 42). Various methods such as contrast test bolus, varying gradient shapes, and k-space sampling have been developed to counteract these problems. Ultimately one needs to be cautious in the interpretation of these different methods of flow data acquisition, as in addition to the aforementioned factors, artifacts or signal loss from the pulsatile, tortuous, and calcified or diseased nature of the vessels can further confound the overall picture (32,33).

Fast MR Imaging

The advantages of fast MRI in the clinical setting are evident to any clinician or radiologist who has dealt with a patient who cannot keep still within the scanner. Various techniques have been developed the most commonly encountered being GRE or fast GRE (Fig. 43), FSE, and EPI. In GRE, the lack of a 180° refocusing pulse is the main difference from a conventional spin echo sequence. Instead of an RF refocusing pulse, the signal is formed by gradient pulse with the MR signal decaying as a function of T2*. In other words, rather than T2

Mri Angiography Artefact

Fig. 41. Normal 3D TOF coronally acquired reconstructed contrast MRA of the arch and neck and most of circle of the Willis (left). High-grade stenosis of the left proximal internal carotid artery and proximal right common carotid artery at the takeoff from the arch, resulting in functionally decreased flow signal of the distal right internal carotid artery in a different patient (right).

Fig. 41. Normal 3D TOF coronally acquired reconstructed contrast MRA of the arch and neck and most of circle of the Willis (left). High-grade stenosis of the left proximal internal carotid artery and proximal right common carotid artery at the takeoff from the arch, resulting in functionally decreased flow signal of the distal right internal carotid artery in a different patient (right).

Mra Neck Care Bolus
Fig. 42. Coronal postcontrast 3D TOF MRA of the neck demonstrating the consequence of missed timing of contrast bolus resulting in an uninterpretable image owing to overlapping of the arterial and venous phases.

3d fast Grad ccho

Fig. 43. Diagram of a 3D fast gradient echo sequence. rf, radiofrequency.

3d fast Grad ccho

Fig. 43. Diagram of a 3D fast gradient echo sequence. rf, radiofrequency.

decay GRE works within a much smaller envelope (Fig. 6): for a T2-weighted GRE image the TE will be 35 ms instead of 80 ms. As illustrated in Fig. 28, a frequency-encoding gradient lobe (with half the area) first dephases the transverse magnetization followed by a second gradient lobe with opposite polarity, thus rewinding the spins phases. In GRE, the RF pulse used is often less than 90° in order to conserve the longitudinal magnetization for subsequent excitation, as the TR is short. This flip angle thus serves as another parameter to optimize the image contrast; having a small angle has essentially the same effect on tissue contrast as decreasing TR. For a Tl-weighted GRE, a larger flip angle (60°) is used, whereas a smaller one is used for T2 weighted imaging (15°). The TE and flip angle therefore also determine tissue contrast, with the exception that when TR is extremely short (30 ms) a steady-state phenomenon comes into play owing to the presence of residual transverse magnetization that is recycled into the subsequent excitation.

To make the scanning even faster, one can further decreases TR (8-10 ms with 5-15° flip angles) or acquire more than one set of data per excitation. As one can predict, with the shorter time comes the penalty of poor SNR, image contrast, and saturation. The shorter time also means that more signal is acquired from pulses early in the echo train than later. This in turn allows for the influence of fc-space sampling on image contrast whereby if the central fc-space is sampled during the stronger signal from the earlier echo train, the resulting image will have better contrast but poorer resolution.

Both FSE and EPI accomplish imaging in a short period by acquiring more datasets per excitation. A basic EPI obtains all the data for an image in a single excitation about 35-50 ms and is essentially a gradient-echo sequence. An RF pulse is turned on at the same time as the slice-select gradient to generate magnetization in a single slice. The frequency gradient is then turned on, alternating in polarity as the echo train is generated (Fig. 44). For the filling of fc-space for each of these echoes, the phase-encoding gradient is turned on very briefly (blipped) starting from the maximum negative value and again with alternat-

Fig. 44. Basic echo planar (EPI) pulse design. rf, radiofrequency.
Fig. 45. (A,B) Diagram of filling of lines in k-space in EPI and FSE sequence consisting of four echo train lengths respectively.

ing polarity for each blip (Fig. 45A). In other words, the whole k-space is filled after a single shot of this echo train; 128 phase-encoding steps will require an echo train of 128 echoes. In order for the echo train to be shorter than T2* decay, the hardware demands stronger and faster gradients than the standard MR fare, resulting in expensive upgrades. Various different modifications of a basic EPI sequence have been developed including variations of k-space filling to save more time but minimizing image degradation.

FSE is in essence the spin-echo version of EPI. Again, it starts with a 90° pulse, but the pulse sequence in FSE is followed by a train of 180° pulses, unlike conventional spin echo (CSE) with its subsequent 90-180° pulses consisting of one phase encoding each. This in turn generates a train of spin echoes each with an independent phase encoding (Fig. 46). For example, in a CSE sequence with four echoes, each echo is phase-encoded in the same way and is stored in separate memory locations with four images generated after reconstruction. In FSE with a four-echo train length (ETL), each echo is phase-encoded differently with all the data placed into the same memory location, thus generating a single

Fig. 46. Basic FSE sequence. rf, radiofrequency.

image (Fig. 45B). The scan time for FSE, however, will only be a fourth that of CSE. To perform independent phase-encoding for each echo, before each readout gradient, a phase encoding gradient imparts phase prep, if you will, to the transverse magnetization. Then immediately after readout, a second gradient of equal and opposite amplitude rewinds it back to zero for the next prep gradient. (Fig. 46). As there are contributions from various TEs, the final contrast of the image is determined by the TE of the echo that fills the center part of fc-space (Fig. 10). The scanner tries to match the echo that is closest to the desired TE (selected by the operator) and maps it to the center of fc-space. On the images, the difference between CSE and FSE includes less susceptibility to metal and blood products in an FSE sequence as well as fat remaining relatively bright on FSE T2 images (34-36).

MR Perfusion

As mentioned above, functional MR is primarily discussed in Chapter 4, although MR perfusion will be touched on briefly here. Unlike MRA or MRV, which essentially deal with bulk flow, MR perfusion focuses more on the tissue level or microscopic blood flow. A contrast agent such as gadolinium DTPA is used in MR perfusion for its T2 (T2*) effects, which causes signal loss in the area of perfusion owing to dephasing of spins from the susceptibility effect of the contrast bolus as it rapidly traverses the capillary beds (37). Relative rather than true cerebral blood volume (CBV) maps are constructed using the tracer kinetic principle, as the arterial input function is not usually measured. Integration of a signal time curve is performed for each voxel: the signal loss is dependent on the contrast concentration and density of vessels per volume of tissue. An actual quantification of CBV can be accomplished by applying arterial input function at a region of interest over a major blood vessel such as the middle cerebral artery. Semiquantitative cerebral blood flow can also be computed by applying deconvolution methods to simultaneous tissue and vessel concentration time curves, with the latter acquired by the indicator dilution theory.

Fig. 47. Perfusion imaging demonstrating a plot of signal loss curve against time. On the left, the bottom curve shows normally perfused tissue and the top curve represents ischemic tissue with decreased and delayed signal loss owing to decreased delivery of contrast agent. The right diagram illustrates the area "under" the signal curve, which is integrated to calculate the relative cerebral blood volume (rCBV).

Fig. 47. Perfusion imaging demonstrating a plot of signal loss curve against time. On the left, the bottom curve shows normally perfused tissue and the top curve represents ischemic tissue with decreased and delayed signal loss owing to decreased delivery of contrast agent. The right diagram illustrates the area "under" the signal curve, which is integrated to calculate the relative cerebral blood volume (rCBV).

For MR perfusion fast MRI techniques are used including gradient echo, EPI, and FSE sequences. The latter has the advantage in tumor perfusion imaging as it is more sensitive to microvasculature, unlike GRE imaging, which often incorporates artifact from surrounding larger vessels. The change in T2 relaxation rate, or dR2 is defined as -ln(S/ S0)/TE, where S is the signal intensity and S0 the baseline signal. This dR2 vs time curve for every pixel is then mathematically integrated to generate the relative (r)CBV map (Fig. 47). The main clinical application of perfusion is in the setting of stroke, in which the dR2 vs time curve can be used to calculate time to peak (TTP), mean transit time (MTT), relative cerebral blood volume (rCBV) and relative cerebral blood flow (rCBF). From these parameters one hopes to derive information or perhaps a threshold whereby viable ischemic tissue can be salvaged (the penumbra region where there is mismatch of diffusion and perfusion). In an area of infarct, one would expect a slower rise or shallower slope of the curve, slower TTP and MTT, and decreased rCBF (38).

One of the major limitations of this MR perfusion technique occurs in areas that are leaky or necrotic or that have extensive breakdown of the blood-brain barrier. The more direct contact of contrast with tissue vs that within capillaries accentuates the T1 effects of gadolinium, thus counteracting T2 effects and resulting in erroneous lower rCBV. This is especially problematic in the tumor setting when one is attempting to derive information on the more vascular or aggressive portion of the tumor or to differentiate tumor growth from treatment changes. Even though numerous methods have been used to correct this problem (such as a corrective algorithm, a presaturation of leaky areas with a small amount of preinjected dose of contrast, use of stronger T2 effect contrast agents,

Magnetization Transfer Contrast
Fig. 48. Theory behind magnetization transfer showing exchange of off-resonant proton with free water proton as a means to suppress background.

and lessening the T1 effect by increasing TR and decreasing flip angle), the leakage problem has still not been successfully eliminated. Other remedies include noncontrast spin-labeled pulse sequences that tag the incoming spins using RF pulse that are subsequently imaged downstream. These spin-labeled inversion recovery EPI sequences carry their own set of drawbacks. Because of limitations from both hardware and software and because of cost-benefit issues, MR perfusion has not been as clinically successful as MR diffusion and is not a routine clinical tool at this time. Further investigation is necessary before MR perfusion can become an accurate means of determining high metabolic areas within tumor or the differentiation of active tumor from treatment changes (39).

Magnetization Transfer

The MT technique is often used to increase the contrast between the background and regions of interest. Very simply, a pulsed or continuous off-resonance low-power RF pulse is applied to saturate the bounded hydrogen protons in proteins and/or macromolecules. Because of the chemical exchange of these bounded protons with free water protons, some of this saturation will be transferred and exchanged by bulk water, thus causing a decrease in MR signal of the water/background (Figs. 48 and 49). The rate of this proton MT can be quantified pixel by pixel (40). In other words, the MT contrast is a reflection of the efficiency of such proton exchange that can be altered in pathological states such as multiple sclerosis in which all imaging sequences may appear normal to the eye but will have abnormal measured MT ratios (41).

Relaxation Audio Sounds Autumn In The Forest

Relaxation Audio Sounds Autumn In The Forest

This is an audio all about guiding you to relaxation. This is a Relaxation Audio Sounds with sounds from Autumn In The Forest.

Get My Free MP3 Audio


Post a comment