MRI-Based Direct Measurements of the T2* Transverse Relaxation Time of Water and Lipid Protons in Water–Lipid Mixtures

Abstract

Magnetic resonance imaging (MRI) has emerged as the method of choice for in vivo quantification of lipids. The MRI methods originally proposed for lipid quantification did not take into account for the loss of signal due to T2*. In the last decade, a number of algorithms has been introduced for the T2* correction. These algorithms assumed that the T2* of the water protons (T2w*) is the same as the T2* of lipid protons (T2L*). More recent algorithms have proposed a more sophisticated model (dual T2* correction), which consider different values for T2w* and T2L*. However, no reference values exist for validating the values of T2w* and T2L* obtained by these algorithms in tissues or phantoms where water and lipid protons co-exist. In the current work, we propose a direct measurement of T2w* and T2L* in phantoms consisting of water–lipid mixtures. MR multi-gradient-echo images were acquired with a low receiver bandwidth/pixel with the aim of separating the water and lipid signals. In all phantoms, T2L* (range 7.6–10.7 ms) was significantly shorter than T2w* (range 48.9–57.4 ms). The proposed method provides the ground truth values of T2w* and T2L* for development, validation, and optimization of lipid quantification methods based on dual T2* correction.