T1 -corrected quantitative chemical shift-encoded MRI

Abstract

Purpose: To develop and validate a T₁ -corrected chemical-shift encoded MRI (CSE-MRI) method to improve noise performance and reduce bias for quantification of tissue proton density fat-fraction (PDFF).

Methods: A variable flip angle (VFA)-CSE-MRI method using joint-fit reconstruction was developed and implemented. In computer simulations and phantom experiments, sources of bias measured using VFA-CSE-MRI were investigated. The effect of tissue T₁ on bias using low flip angle (LFA)-CSE-MRI was also evaluated. The noise performance of VFA-CSE-MRI was compared to LFA-CSE-MRI for liver fat quantification. Finally, a prospective pilot study in patients undergoing gadoxetic acid-enhanced MRI of the liver to evaluate the ability of the proposed method to quantify liver PDFF before and after contrast.

Results: VFA-CSE-MRI was accurate and insensitive to transmit B₁ inhomogeneities in phantom experiments and computer simulations. With high flip angles, phase errors because of RF spoiling required modification of the CSE signal model. For relaxation parameters commonly observed in liver, the joint-fit reconstruction improved the noise performance marginally, compared to LFA-CSE-MRI, but eliminated T₁ -related bias. A total of 25 patients were successfully recruited and analyzed for the pilot study. Strong correlation and good agreement between PDFF measured with VFA-CSE-MRI and LFA-CSE-MRI (pre-contrast) was observed before (R² = 0.97; slope = 0.88, 0.81-0.94 95% confidence interval [CI]; intercept = 1.34, -0.77-1.92 95% CI) and after (R² = 0.93; slope = 0.88, 0.78-0.98 95% CI; intercept = 1.90, 1.01-2.79 95% CI) contrast.

Conclusion: Joint-fit VFA-CSE-MRI is feasible for T₁ -corrected PDFF quantification in liver, is insensitive to B₁ inhomogeneities, and can eliminate T₁ bias, but with only marginal SNR advantage for T₁ values observed in the liver.

Keywords: T1 bias; T1 correction; chemical-shift encoded imaging; fat quantification; hepatic steatosis; liver fat; magnetic resonance imaging; proton density fat-fraction.

Figure 1.

RF spoiling used with SGRE results in near perfect spoiling with of the signal magnitude, but leaves a strong flip angle dependent transverse signal phase. Steady state transverse signal amplitude and phase were calculated using Bloch-equation simulations.

Figure 2.

CRLB analysis can be used to identify optimal flip angle pairs that optimize the SNR performance of the proposed VFA-CSE-MRI method. In these plots SNR is defined as 20/(standard deviation of PDFF estimator). A) Predicted SNR with respect to all flip angle pairs. B) Optimal SNR with flip angle pairs under the constraint of an upper limit. The broad maximum, allows flip angle #1 to be reduced from 33° to 20°with marginal SNR penalty.

Figure 3.

PDFF estimation using VFA-CSE-MRI is insensitive to transmit B₁ inhomogeneities in simulations. In this simulation negligible error in the estimated PDFF was observed. Absolute PDFF error as predicted by simulation in liver fat quantification at 1.5T (A, B) is shown. Note that these simulations assume that the percent error in transmitted B₁ is the same for both flip angles.

Figure 4.

PDFF estimation using VFA-CSE-MRI is insensitive to transmit B₁ inhomogeneities in phantom experiments. Plots show PDFF measured using joint-fit VFA-CSE-MRI in phantoms in the presence of B₁ error. Phantoms were constructed in groups with varying PDFF and T_1W values controlled by doping agent CuSO₄. PDFF measurement with LFA-CSE-MRI (flip angle=1°) was used as the reference.

Figure 5.

Any degree of T₁-weighting leads to bias in PDFF estimation if the T₁ of water and fat are different (A). Simple correction (eg. assuming T_1W = 586ms and T_1F = 343ms), also leaves considerable bias if the true T₁ values are different than assumed values (B). These simulations demonstrate the utility of T₁-corrected methods such as the proposed VFA-CSE-MRI method.

Figure 6.

Noise performance of PDFF estimation using CRLB analysis (solid line) and Monte Carlo simulations (data points), demonstrate that for parameters commonly encountered in the liver that LFA-CSE-MRI methods have the highest SNR performance, although this performance is highly dependent on the flip angle. At very low flip angles (eg. 2°), conventional LFA-CSE-MRI has lower SNR performance. Interestingly, the proposed joint-fit VFA-CSE-MRI shows only slightly improved performance compared to the 2-step VFA method. This is likely due to the need for estimating independent constant phase on the water and fat signals, for the joint-fitting, due to the residual species dependent phase from RF spoiling. Note that SNR is defined as 20/(estimator standard deviation) for each method. The input SNR in these analyses was normalized for acquisition time.

Figure 7.

Modeling for different constant phase values between water and fat resulting from RF spoiling is needed to address the resulting bias in PDFF if this confounder is not considered. This bias can be eliminated in simulations (A) and greatly reduced in phantoms (B). The phantom used for these measurements was that doped with 1mM CuSO₄.

Figure 8.

The proposed VFA-CSE-MRI method eliminates T₁-related bias, as shown in phantom experiments. The degree of bias is highly dependent on the difference in T₁ between water and fat. High flip angle CSE-MRI acquisitions demonstrate large bias, while even low flip angle acquisitions demonstrate measurable bias.

Figure 9.

Example PDFF, R₂* and T_1W maps from a subject with elevated liver PDFF, acquired before and after the administration of gadoxetic acid, visually demonstrating the effects of contrast on estimated PDFF, R₂* and T_1W values. In this figure, the PDFF map and ROI value shown for LFA-CSE-MRI pre-contrast was not corrected with any T₁ assumption.

Figure 10.

Summary results from the pilot clinical study demonstrate strong correlation good agreement between VFA-PDFF and LFA-PDFF before and after contrast, whereas the high flip angle acquisition leads to strong positive T₁-related bias before contrast and strong negative T₁-related bias after contrast. Also shown are R₂* and T_1W before and after gadoxetic acid. A small increase in R₂* is noted and also a strong decrease in T_1W observed, due to the presence of gadolinium. Note one outlier with high T_1W (pre,*) is in a patient with biopsy proven NASH, and a second outlier (post,**) was from a patient with known cholangiocarcinoma and liver failure related to biliary obstruction.