SNR and CNR Comparison of Single-Echo Dixon and Subtraction Contrast-Enhanced MR Angiography

Purpose: Multi-echo Dixon-based methods have recently gained attention for use in contrast-enhanced MR angiography (CE-MRA) due to their ability to suppress the high signal from fat while avoiding subtraction errors due to motion between preand post-contrast images in subtraction CE-MRA. Additionally, multi-echo Dixon CE-MRA provides improved signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) dependent on the number of acquired echoes and the concentration of contrast in the blood. However, multi-echo Dixon-based methods require two or more images acquired at different echo times, possibly extending the repetition time and reducing temporal resolution in time-resolved studies. Single-echo Dixon techniques avoid extending the dynamic scan time by obtaining an estimate of the “nuisance” phase (due to B0 inhomogeneities and the initial phase in the water and fat signals) from either a calibration scan before the dynamic portion of the exam or a “virtual shimming” procedure. Water images from single-echo Dixon are expected to avoid subtraction artifacts and exhibit SNR a factor of √2 higher than that of subtraction CE-MRA. Single-echo Dixon methods have been used for dynamic fat-suppressed exams in the past, but to our knowledge the technique has not been applied to CE-MRA. Moreover, a rigorous SNR comparison with subtraction CE-MRA has not been performed, nor has the effect of gadolinium-induced (Gd-induced) susceptibility on single-echo Dixon SNR been studied at 3T. The purpose of this study is to theoretically derive and experimentally confirm SNR and CNR in single-echo Dixon CE-MRA compared to subtraction CE-MRA.