Comparison of Magnetic Resonance-Based Elastography and Ultrasound Shear Wave Elastography in Patients With Suspicion of Nonalcoholic Fatty Liver Disease

This study investigated the correlation between magnetic resonance elastography (MRE) and shear wave ultrasound elastography (SWE) in patients with clinically diagnosed or suspected nonalcoholic fatty liver disease (NAFLD). Subjects with or at risk of NAFLD identified by magnetic resonance imaging (MRI) proton density fat fraction (PDFF) were prospectively enrolled. For each patient, 6 valid 2-dimensional SWE measurements were acquired using a Logiq E10 scanner (GE HealthCare, Waukesha, WI). A reliability criterion of an interquartile range to median ratio of <= 15% was used for SWE to indicate quality dataset. Magnetic resonance elastography, and MR-fat quantification data were collected the same day as part of the patient's clinical standard of care. Magnetic resonance imaging PDFF was used as a reference to quantify fat with >6.4% indicating NAFLD. Pearson correlation and t-test were performed for statistical analyses. A total of 140 patients were enrolled, 112 of which met SWE reliability measurement criteria. Magnetic resonance elastography and 2-dimensional SWE showed a positive correlation across all study subjects (r = 0.27; P = 0.004). When patients were grouped according to steatosis and fibrosis state, a positive correlation was observed between MRE and SWE in patients with fibrosis (r = 0.30; P = 0.03), without fibrosis (r = 0.27; P = 0.03), and with NAFLD (r = 0.28; P = 0.02). No elastography technique correlated with liver fat quantification (P > 0.52). Magnetic resonance elastography was significantly different between patients with and without fibrosis (P < 0.0001). However, this difference was not apparent with SWE (P = 0.09). In patients with suspected or known NAFLD, MRE, and SWE demonstrated a positive correlation. In addition, these noninvasive imaging modalities may be useful adjunct techniques for monitoring NAFLD.