Magnetic resonance imaging (MRI) techniques can usually be classified into spin-echo (SE) and gradient-echo (GRE) pulse sequences. In this presentation, the basic physical principles of GRE imaging, as well as different mechanisms to generate image contrast will be explained. Differences between SE and GRE MRI will be discussed. Additionally, the influence of different pulse sequence parameters (e.g. echo time, repetition time, flip angle; as well as spoiling techniques and preparation pulses) on the image contrast will be covered. Clinical applications of GRE imaging techniques will be shown exemplarily.
Magnetic resonance imaging (MRI) techniques can usually be classified into spin-echo (SE) and gradient-echo (GRE) pulse sequences. However, also hybrids exist such as the GRASE (gradient and spin echo) technique [1]. The physical principles and clinical applications of GRE imaging have been reviewed in several articles (e.g. by Markl and Leupold [2] and Hargreaves [3]).
Technically, the difference between SE and GRE pulse sequences is the use of refocusing radiofrequency (RF) pulses. In SE pulse sequences, refocusing RF pulses are applied that lead to a re-phasing of the magnetization even in the presence of static magnetic field inhomogeneities. Refocusing RF pulses lead to spin-echoes [4] and result in robust imaging techniques.
In GRE techniques, refocusing pulses are lacking. As a consequence, shorter echo (TE) and repetition times (TR) can be used. TR is the period of time between two excitation pulses. In particular, 3D imaging or other applications where fast data acquisition is required are common applications of GRE imaging. Thus, GRE imaging can usually be regarded as a fast imaging technique. TE is the period of time between the (middle) of the RF pulse and the formation of the gradient-echo. In a motionless sample in a homogeneous magnetic field, the gradient-echo appears at a point in time when the magnetization is in-phase (“coherent”) regardless of its position within the sample along at least one direction [3]. In addition, the lack of refocusing RF pulses in GRE imaging leads to a lower specific absorption rate (SAR), which can be beneficial in high (B0 = 3T) and ultra-high field (B0 ≥ 7T) MRI. Disadvantages of GRE imaging are that these techniques are more prone to off-resonance artifacts. In the presence of magnetic field inhomogeneities, image distortions and signal extinctions can occur. However, the latter can also be exploited for diagnostic purposes, since signal extinctions can also be caused by micro-bleedings. This effect is utilized in T2*- and susceptibility weighted imaging [5].
Signal decay in GRE imaging is usually described by the T2*-relaxation time, whereas in SE imaging, the decay of the transverse magnetization is described by the T2-time. T2* relaxation contains effects of dephasing caused by static inhomogeneities (T2’) and time-varying field fluctuations (T2-relaxation). The relationship between T2 and T2* is described by equation 1. The T2’-decay originates from temporally constant (“static”) magnetic field inhomogeneities. These inhomogeneities can be caused by local differences of the magnetic tissue properties. Dephasing caused by T2’-decay can be fully reversed in SE pulse sequences but is not compensated in GRE imaging. Thus, T2* is always shorter than T2.
$$$\frac{1}{T_2^*}=\frac{1}{T_2}+\frac{1}{T_2'}$$$ (equation 1)
Image contrast in GRE imaging can be generated by a number of mechanisms and techniques. These techniques include the adaptation of the basic pulse sequence parameters TE, TR and flip angle, the application of spoiling techniques and contrast preparation.
TE, TR and flip angle:
Whereas 90° excitation pulses are applied in SE pulse sequences, the flip angles in GRE pulse sequences are usually much smaller than 90°. For example, for a given TR (e.g. 50 ms) and short TE (e.g. 3 ms), image contrast can be changed from a spin-density weighted contrast to a heavily T1-weighted image by solely increasing the flip angle. In this case, small flip angles (e.g. 5°) lead to a spin-density-weighted contrast, whereas a larger flip angle (e.g. > 20°) generates a T1-weighted image. If TR is much longer than T1 (e.g. TR > 5 ∙ T1), the magnetization has sufficient time to recover and, depending on TE, the image is spin-density weighted or T2*-weighted. If TR is comparable to T1 but still much longer than T2*, only partial recovery of the longitudinal magnetization occurs. In this case, the magnetization reaches a steady-state that introduces T1 contrast. However, as mentioned above, the T1 contrast depends also on the flip angle and vanishes for small flip angles.
Spoiling techniques:
In fast gradient imaging, TR is usually short (e.g. TR < 2 ∙ T1 or even TR < 2 ∙ T2*). In this case, both, longitudinal and transverse magnetizations, reach a steady-state and contrast depends on the spoiling techniques and on the flip angle. Image contrast can be a function of T1 and T2* (spoiled GRE) or T2/T1 (balanced SSFP).
The term “spoiling” describes different methods (RF-spoiling or gradient spoiling) that can be applied to cancel (or “to destroy”) transverse magnetization at the end of each TR interval. Gradient and RF spoiling leads to the FLASH-technique [6] (also known as spoiled-GRE (SPGR), T1-fast-field-echo (FFE)). FLASH imaging enables pure T1 contrast and is often used for 3D contrast-enhanced MRI.
In balanced steady-state free precession (balanced SSFP) [7], no spoiling is performed and all gradients are fully rewound (common acronyms are: TrueFISP, FIESTA, Balanced FFE). In balanced SSFP, image contrast is based on T2/T1 and the signal-to-noise ratio (SNR) is usually superior to the SNR of spoiled imaging techniques. This comes at the expense of a higher sensitivity to off-resonance effects. Inhomogeneities of the main magnetic field (“off-resonances”) can cause signal variations and dark bands (“banding artifacts”). An application of the balanced SSFP technique is rapid morphological imaging, especially of the beating heart. The T2/T1-weighted signal generates a good contrast between blood and myocardium. In addition, balanced SSFP sequences are intrinsically flow compensated, which is also beneficial for cardiac MRI [7].
Additional Modification of Image Contrast:
Adaptation of the major pulse sequence parameters TE, TR and flip angle, as well as spoiling already provides a large variety of image contrasts in GRE imaging. In addition, preparation pulses can be used to further optimize the contrast for certain applications. For example, fat saturation pulses (i.e. spectrally selective 90° pulses can be applied to the fat magnetization, followed by a crusher gradient) can be used to suppress signal from fat. Multi-echo techniques are commonly used to acquire images in which fat and water magnetization are in- or out-of-phase. They can also be used in Dixon MRI [8] for fat/water separation and to calculate fat-fraction maps. Additionally, two images with different contrast properties are sometimes combined. This is done, for example, in the fast acquisition double echo (FADE) or double echo at steady state (DESS) sequence [9-11]. These techniques provide strongly T2-weighted images and are often used in cartilage and synovial fluid imaging. Magnetization preparation by an 180° inversion pulse followed by a delay time and a rapid 3D GRE readout is used in magnetization-prepared rapid gradient-echo imaging (3D MP-RAGE) to optimize image contrast [12]. 3D MP-RAGE belongs to the most popular sequences for high-resolution T1-weighted brain MRI. Applications include voxel-based morphometry and contrast-enhanced imaging.