Liver-to-Muscle Signal Intensity Ratio
The GRE-based signal intensity ratio technique is based on observation of a signal intensity decrease from T2* shortening in the presence of liver iron overload(Fig.3,7) in comparison with the signal intensity of a unaffected reference tissue , typically the paraspinal muscles. The signal intensity of normal liver parenchyma should always be higher than that of the paraspinal muscles.A hypointense liver relative to the paraspinal muscles indicates iron overload. This is assessed semi quantitatively by measuring the liver-to-muscle signal intensity ratio. In severe iron overload, the signal intensity of the liver is lower than that of the paraspinal muscles even with sequences with limited sensitivity to iron overload (ie, T1-weighted and PDW). In moderate iron overload, lower signal intensity of the liver relative to that of the paraspinal musclesis seen with sequences that are moderately sensitive to iron overload (ie, T2*-weighted sequences with intermediate TE). In mild iron overload, lower signal intensity of the liver relative to that of the paraspinal muscles is seen with sequences that are most sensitive to iron overload (ie, heavily T2*-weighted sequences with long TE).
The sequence is designed to have a set of five breath-hold GRE sequences with fixed TR and different TE and flip angles, optimised for 0.5, 1.0 and 1.5 Tesla (T) magnetic fields. The method can easily be implemented by virtually all machines in the world. On each sequence, the liver the SI is measured at three ROIs in the right lobe, while muscle SI is measured at ROIs in the right and left paraspinal muscles(Fig 4).
Fig. 3: Pseudonodules (arrows) and iron overload. (a) Ax- ial GRE T2-weighted MR image demonstrates the low signal in- tensity of the liver that is due to iron overload and also shows nodules without expansive effect or vascular distortion. (b, c) Axial GRE T1-weighted out-of-phase MR image (b) demonstrates a decrease in the signal intensity of the nodules compared with the signal intensity of the nodules on the axial GRE T1-weighted inphase MR image (c), a finding suggestive of focal steatosis and spared areas of hemochromatosis. (d) Axial gadolinium-enhanced fat-saturated GRE T1-weighted MR image shows no different enhancement.
References: Marcony Queiroz-Andrade, Roberto Blasbalg et al, MR Imaging Findings of Iron Overload RadioGraphics 2009; 29:1575–1589
Fig. 4: The liver-to-muscle signal intensity ratio technique requires measurement of the signal intensity with three ROIs in the liver (yellow circles), excluding vascular structures, and one ROI in each paraspinal muscle (red circles) with five MR imaging sequences, in which repetition time and flip angle are held constant while varying TE to alter the T1 and T2* weighting. These values are used to compute five different liver-to-muscle signal intensity ratios.
References: Roxanne Labranche, Guillaume Gilbert, Milena Cerny et al, Liver Iron Quantification with MR Imaging: A Primer for Radiologists RadioGraphics 2018; 38:392–412
T2 Relaxometry methods
Relaxometry methods calculate T2 or T2* by fitting a decay models to the average signal intensity atvarious echo times (TEs). These values may also be expressed as relaxation rates R2 (1/T2) or R2* (1/T2*). To obtain reliable measurements of T2 at different levels of iron overload, acquisition sequences with many different TEs are necessary. SI is plotted as a function of echo time and a T2 parametric map is automatically obtained.
R2 relaxometry requires a multisection single spin-echo sequence during free breathing to obtain axial images with repetition time of 2500 msec, TE every 3 msec from 6 msec to 18 msec (at 6, 9, 12, 15, and 18 msec), flip angle of 90°, section thickness of 5 mm, matrix size of 256, and field of view between 350 and 400 mm, depending on the patient’s habitus. R2 measurement for liver iron concentration, is done by selecting an ROI covering the right liver lobe on the largest axial section of the liver.
R2* relaxometry requires GRE sequences(Fig.5) with multiple breath holds to obtain axial images with repetition time of 25 msec, TEs every 0.25 msec from 0.8 msec to 4.8 msec, flip angle of 20°, section thickness of 15 mm, matrix size of 64 3 64 and field of view of 48 3 24 mm, depending on the patient’s habitus, and bandwidth of 83 kHz . More recently, single breath-hold multiecho GRE sequences have been widely adopted, both in research and the clinical setting.
Quantitative Susceptibility Mapping
The presence of a local susceptibility source such as ferritin or hemosiderin leads to an augmentation of the local magnetic field, which has an impact on the measured phase of a GRE sequence. Using an acquisition with two or more echoes, the local magnetic field can be estimated, and the inverse problem relating the measured magnetic field distribution to the underlying susceptibility distribution can be solved. A 3D breath-hold multiecho GRE sequence with appropriate parameters is used.
After calculation of the quantitative susceptibility map, a local relative susceptibility value (ΔB0), generally expressed in parts per million (ppm), can be extracted from an ROI. That susceptibility value can be related to the liver iron content(Fig.6).
Fig. 6: Flowchart of processing steps for quantitative susceptibility mapping.
References: Roxanne Labranche, Guillaume Gilbert, Milena Cerny et al, Liver Iron Quantification with MR Imaging: A Primer for Radiologists RadioGraphics 2018; 38:392–412
Ultrashort-TE Sequences
Ultrashort-TE sequences with TE as low as 0.1 msec are achievable and would permit quantification of massive iron overload, even at 3.0 T. In comparison, the shortest TE achievable with current GRE sequences is in the range of 0.8 msec. Preliminary results suggest that ultra short TE sequences may represent a viable replacement for or alternative to conventional GRE sequences for R2* estimation, for both low and high iron overload.
Multiparametric Assessment
Chronic liver diseases are characterized by the concomitant presence of pathologic changes, including liver fat, iron, inflammation, biliary disease, and fibrosis. Further, the coexistence of several pathologic changes may act as a confounder to liver iron quantification, affecting the R2 or R2* of tissue. Eventually, multiparametric techniques will be required to assess each of these pathologic parameters and control these biologic confounders. Hence, a comprehensive multiparametric quantitative MR imaging protocol may include elastography sequences for assessment of liver fibrosis and multiecho chemical shift–encoded GRE sequences for simultaneous assessment of proton-density fat fraction (PDFF) for liver steatosis quantification and R2* for liver iron quantification within a single breath hold .Proton-density fat fraction is a biomarker that has been validated for accurate and precise assessment of liver steatosis across different systems. Ultimately, combining several parameters may improve classification accuracy and lead to better diagnostic performance than use of single parameters.