It was mentioned in the last chapter that, in contrast to SE, in GRE there may be residual transverse magnetization at the end of each cycle remaining for the next cycle. This residual magnetization reaches a steady-state value after a few cycles and is denoted M_ss.

This residual, steady-state magnetization is added to the transverse magnetization created by the next α radio frequency (RF) pulse and thus increases the length of the vector in the x-y plane (Fig. 21-1). This then yields more T2* weighting. In other words, tissues with a longer T2 have a longer M_ss than do tissues with shorter T2.

Actually, to preserve this steady-state component, an additional step needs to be taken in the pulse sequence. A so-called rewinder gradient is applied in the phase-encoding direction at the end of the cycle to reverse the effects of the phase-encoding gradient applied at the beginning of the cycle (i.e., it “unwinds” the former effect). In other words, the rewinder gradient is nothing but the opposite of the phase-encoding gradient (Fig. 21-2). For instance, if gradient +3 is applied for phase encoding, the rewinding gradient would be −3.

The word “spoiling” refers to the elimination or “spoiling” of the steady-state transverse
magnetization. There are various ways of accomplishing this task:

RF Spoiling. RF spoiling is the method used in SPGR and is illustrated in Figure 21-3. In this scheme, a phase offset is added to each successive RF pulse. This causes a corresponding phase shift in successive M_ss vectors. By maintaining a constant phase relationship between the transmitter and the receiver (achieved via a phase-locked circuit), successive M_ss vectors cancel each other out. A pulse sequence diagram (PSD) for SPGR is shown in Figure 21-4. In this scheme, rewinder gradients are naturally not used because their purpose is to preserve the steady-state magnetization, thus defeating the purpose of spoiling. An example is seen in Figure 21-5.

Variable Gradient Spoilers. Spoiling can also be achieved by using gradient spoilers. This is accomplished by introducing an additional gradient with variable strengths from cycle to cycle (Fig. 21-6).

Lengthening TR. The last method to achieve spoiling of M_ss is by lengthening TR. When TR is sufficiently large (generally over 200 msec), there is enough time to allow complete dephasing of the spins in the transverse plane (because TR » T2*). This is similar to the SE pulse sequence.

Question: For a long TR (e.g., 500 msec), what is the difference between GRASS (or FISP) and SPGR (or FLASH)?

Answer: There isn’t any! A TR of 500 msec in GRASS effectively allows transverse magnetization to decay away over each cycle, that is, it eliminates the steady-state component (M_ss). Consequently, GRASS/FISP and SPGR/FLASH would have similar properties for any given TE and α when TR is long.

SPGR/FLASH Tissue Contrast. By eliminating the steady-state component, only the longitudinal component affects the signal in the SPGR/FLASH technique. Thus, this technique lends itself to reduced T2* weighting and increased T1 weighting. This is true provided α is also relatively large. When α is small, however, the T1 recovery curves play a minor role, and proton density (PD) weighting is increased.

This technique is harder to comprehend. It yields heavily T2 (not T2*)-weighted images. The PSD is shown in Figure 21-7. The idea here is that each α RF pulse contains some 90° and some 180° pulses embedded in it. Therefore, in Figure 21-7, α₁ acts like a 90° excitation pulse and α₂ like a 180° rephasing pulse. This yields a pulse sequence similar to spin echo (SE) whereby an echo is formed at α₃. Because it is difficult to read a signal and transmit α₃ at the same time, the echo is actually recalled 9 msec prior to α₃ by using an appropriate gradient. Note that the echo corresponding to α₁ is formed between α₂ and α₃. Interestingly, in this scheme, TE is larger than TR (and smaller than 2TR by 9 msec), which is somewhat counterintuitive. The rewinder gradients are also shown in the diagram. The rewinder gradient is one cycle ahead of the phase-encoding gradient due to the mechanism described in the foregoing discussion. (In this technique, any two successive RF pulses can be employed to create an SE.)