Hepatic MRI for fat quantitation: its relationship to fat morphology, diagnosis, and ultrasound.
PURPOSE: The value of MRI and ultrasound in quantifying hepatic steatosis is assessed and the results compared with those obtained by liver biopsies. METHODS: A total of 38 patients undergoing hepatic biopsy for a variety of liver diseases were recruited for this study. Hepatic fat morphology and severity were assessed visually in each biopsy specimen. Steatosis pattern included macrovesicular, microvesicular, or mixed. The severity of hepatic steatosis was assessed by MRI through chemical shift imaging (n = 38) and by ultrasound through echogenicity (n = 31). RESULTS: MRI had a better correlation than ultrasound for microscopic fat content (r = 0.77, P < 0.001 vs. r = 0.41, P < 0.05). In macrovesicular steatosis, MRI and ultrasound both correlated well with microscopic fat content (r = 0.92, P < 0.001 vs. r = 0.90, P < 0.001). In nonalcoholic fatty liver disease, ultrasound revealed severe steatosis in all instances, but MRI fat content ranged greatly (19%-40%). In diagnoses excluding nonalcoholic fatty liver disease, increasing ultrasound severity did not correspond to advanced MRI fat content. CONCLUSION: Hepatic MRI and ultrasound are both useful in identifying heavy fat accumulation associated with nonalcoholic fatty liver disease. MRI is superior to ultrasound in detecting and quantifying minor degrees of fatty metamorphosis in the liver.