Ronald Ouwerkerk PhD, Presenter: Nothing to Disclose

Ranganath Muniyappa MD, Abstract Co-Author: Nothing to Disclose

Christopher E. Ramsden MD, Abstract Co-Author: Nothing to Disclose

Monica C. Skarulis MD, Abstract Co-Author: Nothing to Disclose

Ahmed Medhat Gharib MBChB, Abstract Co-Author: Nothing to Disclose

To test and validate a new 1H-MRS method to quickly and accurately measure hepatic lipid content, capable of measuring with low fat signals or small changes in fat content.

Localized MRS (8ml volume) was optimized for signal-to-noise efficiency of fat to obtain T1, T2 and spin density for water and lipids in human livers within one breath-hold. After recording an initial fully relaxed scan, spectra were acquired at a short repetition time (TR) in steady state with varying TE values and 4-fold signal averaging. Water and lipid T1 were estimated from the fully relaxed and steady state signal ratios and used to correct T2 measurements. T2 and spin density were obtained from linearized regression fits of exponential decay with TE. T1 and T2 corrected spin densities of water and fat (CH2+CH3) were used to calculate fat fractions (ff) as fat/(fat+water). Thirteen healthy subjects were assessed with this fast TR method (FTR) and with HISTO, an established method for single breath-hold T2 corrected ff measurement at 3s TR. Longer TR avoids T1 correction but only one scan per TE value can be recorded. Both methods were compared with ff determined by much lengthier respiratory navigator-gated scans (nav-MRS).

The figure A) shows spectra recorded with HISTO, last spectrum (TE 72 ms) shown with scale x50 and B) spectra recorded with FTR from the same volume. The fully relaxed spectrum shown at 1/2 scale (dashed) and lipid signal at TE 144 ms at 50x scale. The new method measurably improved sensitivity and accuracy of the liver fat measurement: the liver ff determined with FTR show a better correlation with nav-MRS results (with ff ranging from 0.2% to 7.8%) than HISTO (correlation coefficient R2=0.996 vs. 0.896 with HISTO, N=13). Also, the quality of the T2 regression fit was much better for FTR than HISTO as evidenced by much better linearized regression fit coefficient R2 for fat signals: FTR R2 = 0.951± 0.053 vs. HISTO 0.760±0.271, N=13.

Signal averaging and more sampled TE values result in more sensitive and accurate estimates of T2 corrected fat fractions, particularly in livers with low lipid content.

This MRS method for measuring the liver fat content has the ability to detect lower ff and smaller changes in ff than current water-fat imaging methods which makes it a particularly useful tool for monitoring disease progression or the effects of therapy.

Ouwerkerk, R, Muniyappa, R, Ramsden, C, Skarulis, M, Gharib, A, An Efficient and Sensitive 1H-MR Spectroscopy Method for Quantifying and Monitoring Hepatic Steatosis with T1 and T2 Corrected Fat Fractions.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14015114.html