Advanced MRI techniques estimate proton density fat fraction (PDFF) by acquiring multi-echo, gradient-echo images to correct for R2*, with low flip angle and long TR to minimize T1 weighting. Images are analyzed with a multi-peak fat spectral model to correct for multi-frequency interference effects. These techniques assume that R2* of fat and water are identical and therefore measure and correct for a single R2* value. Previous MRI studies suggested that this single R2* value increases as PDFF increases¹, but these studies did not examine the relationships between PDFF and the separate R2* values of fat and water. The purpose of the present study was to examine the relationships between PDFF and the separate R2* values of fat and water in adults with fatty liver disease. We used MRS to measure R2* of water and fat, since MRS permits reliable measurement of these values, while MRI estimates may be unstable.

Figure 1 shows fat peaks are not simple singlets, but have complex structure due to j-coupling. The fat spectral model used by MRI and MRS to estimate PDFF in human liver in vivo neglects this complexity and instead assumes each peak is a broad singlet. This assumption causes the effective R2* (R2*^eff) of fat to be greater than the true R2* of fat, since R2* is correlated to peak-width. In this work, we use MRS to measure fat R2*^eff, rather than true fat R2*, since fat R2*^eff is the relevant parameter for PDFF techniques. Water is a single peak uncomplicated by j-coupling; hence, for water, R2* and R2*^eff are the same.

¹H STEAM MR spectra were acquired at 3 Tesla (GE Signa EXCITE HDxt, GE Healthcare, Waukesha, WI) using an 8-channel torso array coil in 46 adults (22 male, 24 female; ages 24-71 yrs, mean 49.3 yrs) with fatty liver disease and MRS-determined PDFF > 5%. Subjects with PDFF < 5% were excluded as fat R2*^eff cannot be measured well when PDFF is less than 5%. After conventional imaging, a 20x20x20 mm voxel was selected within the liver that avoided liver edges and large biliary or vascular structures. Following a single pre-acquisition excitation, five spectra (TR 3,500 ms, TM 5 ms) were acquired with a single average at progressively longer TEs of 10, 15, 20, 25 and 30 ms in a single 21 s breath-hold.

Spectra from individual channels were combined using singular value decomposition². A single experienced observer analyzed the spectra using the AMARES algorithm³ included in the MRUI software package⁴. Fat spectra were fitted with two different sets of prior knowledge. To calculate PDFF, spectra were analyzed with standard established prior knowledge⁵ in to give T2-corrected peaks area of water (4-6 ppm) and fat (0-3 ppm). Spectra were then corrected for fat included in the water peak using previously-determined liver spectra⁶, allowing PDFF to be calculated. Spectra were also analyzed using prior knowledge based on the fat spectrum used in many MRI techniques⁶. The fat spectrum was modeled with nine Gaussians with locations fixed relative to each other and with identical peak-width (i.e., each peak was assumed to have the same R2*^eff). The water peak was modeled by a single unconstrained Gaussian. The areas of the fat peaks were left unconstrained as the fat peak areas may not match those in reference spectrum⁶ after allowing for T2 decay. Water R2* and fat R2*^eff were calculated for each TE, and the average values recorded. Linear regression analyses were performed.

Figure 2 compares water R2* and fat R2*^eff with PDFF. There is no evidence of a change in water R2* with PDFF (R² 0.006). There is a weak trend for fat R2*^eff to decrease with increasing PDFF (slope -0.80, intercept 85.36 s^-1, R²: 0.25). Figure 3 shows the difference between water R2* and fat R2*^eff. At low PDFF values water R2* is greater than fat R2*^eff, but at higher PDFF values fat R2*^eff is greater than water R2*.

Water R2* is independent of PDFF, whereas fat R2*^eff decreases slightly as PDFF increases. Water R2* is lower than fat R2*^eff at low PDFF and higher than fat R2*^eff at high PDFF. These findings do not completely explain the increase in R2* with PDFF observed in MRI, and further research is needed to understand these unexpected results.