Spin echo sequences are the classical “working horses” of MRI.

Sequences with a short echo time TE (minimal TE, about 15 ms) and a short TR (400–600 ms) are influenced mainly by T1 relaxation. Short TE, long TR gives a proton-density image. Long TE, long TR results in images influenced more by T2 relaxation.

The shorter T1 time at lower-field strength provides higher signal at short TE and a better T1 contrast.

Minimum TE is depending on the gradient performance in terms of gradient amplitude and, more importantly, gradient slope (or rise time). The combination of these two parameters is called “slew rate.”

Some younger low-field scanners have improved gradient performance.

A good example is gradients with 24 mT/m amplitude and a rise time of 450 ms. The slow rate is therefore 55 T/m/s, enabling sufficiently short echo time.

Since T1 relaxation is faster, the time until recovery of longitudinal relaxation is completed is shorter: TR can be reduced to <500 ms, providing some gain in scan time, at equal signal-to-noise and contrast-to-noise ratio (SNR and CNR).

T2-weighted sequences do not take profit from lower field strength, contrast-enhanced T1 sequences likewise, due to the reduced contrast effect of Gd (see below).

To increase scan speed, one excitation pulse can be followed by multiple inversion pulses (multi-echo). The higher the number of echoes/TR, the faster is the sequence. Since TR is short in T1 sequences, they take only limited profit. For T2 sequences with long TR, the number of echoes can go up to 100/TR and above. The late echoes, however, have lost most of their signal (Fig. 6.2).

To increase signal strength for sequences with long echo trains, a trick is played: The early echoes fill the central part of the signal space (k-space) defining image contrast. The late echoes fill the peripheral part of the k-space, giving mainly contour information (Fig. 6.3).

Peter Rinck gives a beautiful example for the k-space function (Rinck 2009) (Fig. 6.4)

In a gradient echo sequence, spin rephasing is not achieved using an inversion 180° RF pulse, but by inverting the frequency encoding gradient. Keep it simple! (Fig. 6.5)

Since there is no time needed for an echo refocusing pulse, TE can be shorter, and therefore, gradient echo sequences are better suited for shorter T1 times in low-field imaging.

A big drawback of gradient echo is the strong influence of gradient field inhomogeneities, particularly in T2 images. These are compensated, if the spins “go back the same way they came” after a refocusing RF pulse. Therefore, T2-weighted images in gradient echo sequences are addressed to as T2* images.

Further inhomogeneities occur through susceptibility effects, causing areas of rapid dephasing, black ribbons, on bone–fat or bone–air borders (paranasal sinus). To reduce this additional T2* effect, short T1 times are preferred, again supporting the advantages of low-field scanners.

Susceptibility effects are used to detect small iron deposits in areas of micro-hemorrhage (sequence names: GE, SWAN; Siemens, HEMO).

One way to increase the scan speed of gradient echo sequences is to reduce the proton flip angle. The excitation pulse is selected <90°. Therefore, only a part of the protons is flipped. Since TR is short, the residual z-magnetization increases the signal for the next excitation pulse. Haase and coworkers invented this technique in 1986 and called it FLASH (fast low-angle shot) (Haase et al. 1986). The optimal “flip angle,” yielding maximum signal, is called the Ernst–Winkel.
Therefore, with shortened T1 times in low-field imaging, larger flip angles are possible (Fig. 6.6).

In the late 1970s, Peter Mansfield had a brilliant idea (Mansfield and Maudsley 1977). A single excitation pulse, followed by a series of strong gradients, resulting in a series of gradient echoes (echo train) (Fig. 6.7).

This extremely fast imaging is strongly influenced by susceptibility effects. Today, it is mainly used for diffusion-weighted imaging (Fig. 6.7).

EPI imaging takes advantage from higher field strength and gradient performance. Therefore, on low-field scanners, it is possible, but less frequently used. There are concerns about safety of EPI imaging: The rapid gradients can cause eddy currents, potentially leading to neuromuscular stimulation.

If multiple gradient echoes are acquired (fast gradient echo), a further increase of speed is possible. The problem is incomplete recovery of magnetization. Different approaches are realized:

Ultrafast gradient echo uses extremely short TE (2 ms) and TR (3–5 ms); the flip angle is about 5°, giving poor tissue signal. Acquisition time is <1 s.

A 180° inversion pulse starts the sequence, giving the option to use the zero passing and selectively suppress fat or water or silicone (breast imaging).

The readout module may use a single shot (excitation) with variable flip angle or multiple segments (multi-shot). If multiple lines are acquired after a single pulse, the pulse sequence is a type of gradient echo planar imaging (EPI) pulse sequence (Fig. 6.8).

The first MRI examinations in the late 1970s and early 1980s were three-dimensional sequences. The larger the excited volume, the smaller the noise component. Therefore, 3D imaging has a much better SNR than 2D imaging. A 3D sequence is produced by applying an RF excitation pulse without slice selection gradient to a regular 2D sequence, exciting the whole imaging volume.

The disadvantage is the long examination time, which made gradient echo sequences the typical 3D technique.

In 2007, Erik Schweitzer made a far seeing statement during David Stoller’s course on musculoskeletal MRI. He said that in maybe 10 years, an MRI could be performed with only one 3D turbo spin echo sequence. It didn’t take 10 years, but it was a close guess.

A regular fast (turbo) spin echo sequence with an echo train (turbo factor) of 20 and 256 × 256 matrix and 128 slices (the interlacing slice is interpolated to get an isometric volume) takes at a TR of 2.5 s 68 min. Using parallel imaging, this can be shortened to 34 min. Still too long for clinical routine.

The newest 3D TSE sequences (Dr. Schweitzer was right) are called SHAPE in the Siemens world, VISTA on Philips, and CUBE on GE machines. They have an echo train length of >100. With such a long ETL, there is usually only very few signal left for the late echoes. This signal has to be increased.

A possible trick is a refocussing pulse with variable reduced flip angle. Flip angles below 180° lead to reduced early echo intensities but conserve more signal for the late echoes. The original name was FSE-XETA (fast spin echo with extended echo train acquisition) (Fig. 6.9).

Another trick is a three-dimensional reconstruction kernel for parallel imaging using self-calibration and a three-dimensional interpolation of missing data. The algorithm is called ARC (Autocalibrating Reconstruction for Cartesian imaging) (Fig. 6.10).

The resulting 3D TSE sequence is able to scan a T2-weighted sequence with a TR of 2.5 s, an ETL of 100, and a 256 × 256 × 256 volume in 4:30 min. The RF power is low, with metal artifacts reduced. SNR and tissue contrast are spin echo like.

These sequences can be performed on low-field scanners with a to some extent longer scan time (due to less SNR).

Fat is bright in T1, T2, and proton-density imaging. Contrast enhancement results in signal increase in pathologic (inflammatoric, neoplastic, hyperperfused) areas. Edema is bright in T2-weighted sequences. To increase detectability of these pathological findings, the suppression of bright fat signal is mandatory.

Different concepts and techniques exist for fat suppression. Since field strength has an important effect on fat suppression effects, they have to be briefly addressed.

There are two principles for fat suppression: relaxation dependent (STIR) and chemical shift dependent (Dixon, spectral saturation, water excitation, and SPAIR) (Horger and Kiefer 2011).

Diffusion imaging uses the movement of water molecules as contrast-giving principle. Different sequence types are used.

Basically, it is a T2-weighted spin echo, gradient echo, or EPI sequence, using a 180° – rephasing pulse surrounded by a bipolar gradient pair of equal strength. This means that movements between this gradient pair are not completely rephased, which leads to a signal loss. The signal loss is proportional to the spin movement, which in turn depends on the presence of diffusion barriers, like cell membranes. In case of disease (tumor, ischemia), the cell membranes can be damaged, which leads to increased diffusion.

sensitivity of measurement concerning diffusion is based on strength and duration of the gradient pair, as well as the time interval between the gradient activation. These conditions are summarized in a “sensitivity coefficient” b, given by
with γ = gyromagnetic constant of hydrogen, G = strength of diffusion gradient, δ = duration of the gradient, and Δ = time interval between the gradient. The higher the value, the higher the signal loss by diffusion of water molecules.