Background Gradient Correction for Water and Fat Quantification in 2D Liver Imaging

Introduction Fatty infiltration of the liver is one of the primary features of nonalcoholic fatty liver disease. Therefore, accurate quantification of the liver fat content is an important factor in detecting hepatic diseases. MRI is an efficient noninvasive tool for assessing liver fat content. A commonly used liver fat quantification technique with MRI is a dual echo imaging (two gradient sequences with different echo times). This technique relies on the chemical shift between water and fat. However, a more accurate approach is required to correct for T2* shortening effect such as iron deposition. In this matter, a multi-echo gradient echo imaging method can be used to compensate the T2* shortenings [1,2]. When using a multi-echo approach, the presence of macroscopic field inhomogeneities also shortens T2* values and can lead to underestimated T2* values of fat content in liver. Here, we propose to correct for these macroscopic inhomogeneities to accurately quantify T2* values and fat, water content using multi-echo 2D liver imaging. We base on our observation that for 2D imaging, inhomogeneities are dominated in the slice direction by a background gradient component [3].