Flow MRA and MRV

One of the main advantages of MR over other imaging modalities is the ability to evaluate flow (blood or CSF) in a noninvasive manner with or without the need for contrast agents. Moving spins exhibit two properties that are fundamental to the understanding of flow effects in MRI. Time of flight describes the position change of spins as they move in and out of the imaging volume (slice or pixel) during the pulse sequence, i.e., the effect of time elapsed between RF labeling and sampling of moving spin magnetization (Fig. 27). Time-of-flight effects can result in either signal increase or decrease depending on the pulse sequence's repetition time, the T1 of flowing fluid and stationary tissue, the velocity of the fluid, and slice thickness. The aim is to maximize the signal of the flowing spins and minimize the stationary background tissue signal.

When spins within the voxel are repeatedly excited and the TR (time between two 90° pulse) is shorter than the T1 of stationary tissue, there will be very little contribution of tissue signal, as there is not sufficient time for longitudinal relaxation and recovery of magnetization vector for the next RF pulse. In contrast, for the fully relaxed or unsaturated inflowing spins that have not been excited and are moving into the volume between 180° and the next 90° RF pulse, much signal will be generated. The proportion of unsaturated vs satu-

Fig. 27. In the time-of-flight (TOF) technique, the maximum signal results from an inflowing velocity that completely replaces the imaging slice with fresh or unsaturated spins.

_^_readout

Grad echo

Fig. 28. A basic gradient-echo pulse sequence design.

rated spins will then ultimately determine the final average signal within the pixel volume. The relationship among the slice thickness, TR, and velocity is described by the formula ^max = slice thickness/TR where ^max represents the maximum inflowing velocity that completely replaces the slice with fresh spins at TR and therefore generates the most signal. A velocity too low or a slice too thick will result in a higher proportion of saturated spins and therefore decreased overall signal. For a velocity greater than Vmax, there will be a loss of signals, as more unsaturated spins are exiting than entering the slice between the 90° and 180° RF pulse. This is particularly important in spin-echo sequences: both 90° and 180° RF pulse are slice selective, and a 90° "pulsed" spin that exits the slice before the 180° RF pulse will therefore not give a signal. In addition, a longer TE will also favor signal loss as the spins are more likely to have left the imaging volume (27,28).

In contrast to spin echo, gradient-echo sequences (Fig. 28) operate within the T2* envelope where the signal is generated by gradient reversal that is non-slice selective. Any spin that has been excited during the initial slice or slab excitation will give a signal even after it has traveled out of the imaging volume. In z z

Fig. 29. (A-C) In gradient echo, the size of the flip angle affects the degree of longitudinal magnetization, loss being greater with larger angles. The resulting partially recovered magnetization vector then becomes the "starting" vector for the next pulse. This is particularly applicable in the steady-state gradient-echo sequence.

Fig. 29. (A-C) In gradient echo, the size of the flip angle affects the degree of longitudinal magnetization, loss being greater with larger angles. The resulting partially recovered magnetization vector then becomes the "starting" vector for the next pulse. This is particularly applicable in the steady-state gradient-echo sequence.

other words, gradient-echo imaging maximizes the high signal of incoming spins while minimizing the outflow signal loss. The flip angle (a) is important in gradient-echo imaging, in which a larger angle contributes to more effective saturation of stationary spins, especially those with longer T1 magnetization (Fig. 29). One should also note that when flow velocity is not constant or cyclic, there would be loss (and less often gain) in signal owing to turbulence in both CSF (Fig. 30) and vascular flow.

Besides TOF, the other property that is important in understanding flow in MRI is phase shift effects owing to the phase changes that the spin experiences as it moves through magnetic gradients. There can be signal loss caused by cancellation of signal from phase dispersion from different spin velocities, signal increase at even-numbered echoes, and artifacts such as ghosting caused by misregistration of phase shifts from velocity changes. Normally, in stationary tissue the net phase shift caused by the gradients (except the phase-encoding direction) is zero, as the gradients are symmetrical in strength and duration and spins are refocused by the 180° RF pulse in spinecho sequences. When flowing spins have the same velocity, the phase shift is uniform within the voxel, and signals actually add together constructively. Variable velocities owing to acceleration or direction change produce a net signal loss caused by phase spread or dispersion. There can also be a temporal variation such as in the respiratory and cardiac cycle, in which more rapid flow in systole causes more intravoxel dephasing and therefore more signal loss. One can almost assume that in turbulent or complex flow there will be

Fig. 30. CSF flow artifact along the posterior aspect of the T-spine on T2 images manifested as multiple amorphous areas of decreased signal within the CSF that is not substantiated on T1 images (left). Unilateral and often bilateral increased or decreased oval area of signal abnormality within the lateral ventricles caused by flow artifact from CSF traversing the foramen of Monro (right).

Fig. 30. CSF flow artifact along the posterior aspect of the T-spine on T2 images manifested as multiple amorphous areas of decreased signal within the CSF that is not substantiated on T1 images (left). Unilateral and often bilateral increased or decreased oval area of signal abnormality within the lateral ventricles caused by flow artifact from CSF traversing the foramen of Monro (right).

lack of signal from phase dispersion, often occurring at sites such as the carotid bulb, the siphon, or sites of vessel tortuosity (Fig. 31). Various methods have been employed to counteract such higher order motion including gradient moment nulling, creating different gradient waveforms and timing, and employing respiratory and cardiac gating; these techniques are beyond the scope of this chapter (27,29).

MRA and MRV are made possible by means of exploitation of the TOF and phase shift properties mentioned above. The TOF method is based on longitudinal magnetization and in and out flow effects; the other method is the phase-encoding technique. In TOF, we have already discussed that the TR is kept short to minimize the background signal although tissues with very short T1 such as fat or methemoglobin in subacute hemorrhage will cause "shine-through" (Fig. 32). This is often noted on TOF MRA source images, in which the orbital and subcutaneous fat appears bright and is "cut out" on postprocessing images to give the illusion that only the vessels of interest were imaged (Fig. 33). It is thus imperative to note that the TOF technique is really a measure of signal from spins with short T1 or unsaturated spins and is therefore an indirect measurement of flow. This is in contrast to the phase-encoding technique, which

Indirect Tof
Fig. 31. Bilateral subtle loss of flow signal at the carotid bulb and petrous carotid on TOF MRA owing to flow turbulence.
Indirect Tof
Fig. 32. 3D TOF image of the circle of Willis shows "shine-through" of subcutaneous and intraorbital fat as well as a large right subdural hematoma owing to the short T1 effects of fat and subacute blood.
Phase Contrast Mrv
Fig. 34. A simplified version of a 3D phase-contrast MRA pulse sequence. rf, radiofre-quency.

shares the same basis as the previously mentioned technique of phase encoding along the y-axis. The only difference is that instead of position data acquired along the y-direction, velocity is encoded. This is achieved by applying two equal and opposite sign gradient pulses (bipolar pulses) whereby the phase shift in stationary tissue is effectively cancelled out and only moving spins will show a phase shift proportional to the distance traveled (Fig. 34). In other words, a single gradient pulse measures the spin position at the center of the pulse, and a bipolar gradient measures the distance traveled between the centers over time. The major drawback of this technique, which actually measures

Was this article helpful?

0 0
Reducing Blood Pressure Naturally

Reducing Blood Pressure Naturally

Do You Suffer From High Blood Pressure? Do You Feel Like This Silent Killer Might Be Stalking You? Have you been diagnosed or pre-hypertension and hypertension? Then JOIN THE CROWD Nearly 1 in 3 adults in the United States suffer from High Blood Pressure and only 1 in 3 adults are actually aware that they have it.

Get My Free Ebook


Post a comment