Fat suppression is one of the most important techniques in Magnetic Resonance Imaging (MRI) because it improves the visualization of lesions (figure 1), helps in the evaluation of fat content in tumours and better defines the pathology after the administration of paramagnetic contrast agent. At the same time, it greatly improves the signal- and contrast-to-noise ratio (SNR & CNR) and avoids chemical shift artifacts, especially at higher magnetic fields (≥ 3.0 Tesla).
Figure 1 Coronal T1-w SE (left) and fat-suppressed PD-w FSE (right) MR images of the right knee. Fat-suppressed PD-w image can easily visualize the bone marrow edema on the medial condyle of tibia, while the lesion is not well highlighted on the T1-w image. Image dataset was acquired at 1.5 Tesla.
Fat suppression can be achieved by many techniques. Most of these techniques rely on the chemical shift (chemical/spectral saturation, Dixon, water excitation). The mentioned techniques use as an advantage the different precessional frequency between the water and fat protons (the chemical shift between fat and water protons is equal to 3.4 ppm). On the contrary, STIR sequence uses the different T1 relaxation times of the fat and water. Finally, there are hybrid techniques (SPIR, SPAIR) which combine both of these features.
Chemical (Spectral) saturation technique
This technique is based on the chemical shift. The application of a narrow band frequency selective RF pulse (90°) excites the fat protons. The transversal magnetization is destroyed afterwards by spoiler gradients, thus no fat magnetization is left for imaging.
The Dixon technique is based on the chemical shift. With this technique two series of images are acquired. In the first series the signal from fat (Sf) and water (Sw) protons are "in-phase" (Sin). In the second series the signal from fat and water protons are "opposed-phase" (Sop). A separate fat and water image can then be calculated from the equation:
Sin = Sw + Sf
Sop = Sw – Sf
Sin + Sop = [(Sw + Sf) + (Sw - Sf)] /2 = (2 Sw) /2 = Sw
Sin - Sop = [(Sw + Sf) - (Sw - Sf)] /2 = (2 Sf) /2 = Sf
Dixon delivers up to 4 contrasts in one sequence: in-phase, opposed-phase, water and fat images (figure 2).
Figure 2 Three-dimensional (3D) T1-w GRE Dixon images. Dixon technique has the ability to deliver up to 4 contrasts in one measurement. In-phase (upper left), opposed-phase (upper right), water (bottom left) and fat (bottom right) images of the liver. Image dataset was acquired at 1.5 Tesla.
This technique is also based on the chemical shift. A binomial (or composite) excitation pulse is used, which achieves minimum excitation of fat protons and maximum excitation of water protons. Thus, the fat protons will not contribute to the MR signal.
STIR (Short Tau Inversion Recovery)
STIR sequence is based on the different T1 relaxation times of water and fat protons. Fat has a much shorter T1 relaxation time than water and other tissues. Prior to the excitation pulse of the sequence an inversion pulse (180°) is applied, which inverts the spins of all tissues.
When choosing TI such that the longitudinal magnetization of fat at the time when the excitation pulse is applied is zero, the fat protons will not contribute to the MR signal. TI value should be in the range of 130-180 ms at 1.5 Tesla and in the range of 190-240 ms at 3.0 Tesla. By modulating TI, it is possible to adjust the strength of the desired fat suppression (figure 3).
Figure 3 Effects of STIR imaging in hips. By modulating TI, it is possible to adjust the strength of the desired fat suppression. The most powerful fat suppression of the bone marrow is achieved with a TI value of 180 ms (bottom left), which is close to the exact nulling time of the fat tissue. A little lower TI value (160 ms, upper right image) provides the most powerful fat suppression of the subcutaneous fat. Image dataset was acquired at 1.5 Tesla. Images courtesy of Bac Nguyen
Hybrid (SPIR, SPAIR)
Hybrid techniques for fat suppression are a combination of chemical presaturation technique and STIR sequence.