A paradoxical signal intensity increase in fatty livers using opposed-phase gradient echo imaging with...

With the increase in obese and overweight children, nonalcoholic fatty liver disease has become more prevalent in the pediatric population. Appreciating subtleties of magnetic resonance (MR) signal intensity behavior from fatty livers under different imaging conditions thus becomes important to pediatric radiologists. We report an initially confusing signal behavior-increased signal from fatty livers when fat-suppression pulses are applied in an opposed-phase gradient echo imaging sequence-and seek to explain the physical mechanisms for this paradoxical signal intensity behavior. Abdominal MR imaging at 3 T with a 3-D volumetric interpolated breath-hold (VIBE) sequence in the opposed-phase condition (TR/TE 3.3/1.3 ms) was performed in five obese boys (14+/-2 years of age, body mass index >95th percentile for age and sex) with spectroscopically confirmed fatty livers. Two VIBE acquisitions were performed, one with and one without the use of chemical shift selective (CHESS) pulse fat suppression. The ratios of fat-suppressed over non-fat-suppressed signal intensities were assessed in regions-of-interest (ROIs) in five tissues: subcutaneous fat, liver, vertebral marrow, muscle and spleen. The boys had spectroscopically estimated hepatic fat levels between 17% and 48%. CHESS pulse fat suppression decreased subcutaneous fat signals dramatically, by more than 85% within regions of optimal fat suppression. Fatty liver signals, in contrast, were elevated by an average of 87% with CHESS pulse fat suppression. Vertebral marrow signal was also significantly elevated with CHESS pulse fat suppression, while spleen and muscle signals demonstrated only small signal increases on the order of 10%. We demonstrated that CHESS pulse fat suppression actually increases the signal intensity from fatty livers in opposed-phase gradient echo imaging conditions. The increase can be attributed to suppression of one partner of the opposed-phase pair that normally contributes to the destructive interference between water and fat. The result is a paradoxical increase in signal from fatty liver that will depend on both fat content and the relative longitudinal relaxation times of fat methylene protons and water.