Comparison of cardiac MRI on 1.5 and 3.0 Tesla clinical whole body systems

Abstract

Rationale and Objectives: A cardiac imaging pilot study was performed on 1.5 and 3.0 Tesla (T) whole body magnetic resonance units equipped with identical gradient sets and geometrically equivalent body coils. The goals were to compare the signal-to-noise (SNR) and contrast-to-noise (CNR) ratios on matched studies conducted at both field strengths and demonstrate the potential for functional and morphologic cardiac evaluation at 3.0 T. Methods: Short axis cine true fast imaging with steady precession (True FISP) was compared at 1.5 and 3.0 T using the body coil in transmit-receive mode and transmit-only with single loop and phased array receiver coils. SNR of the myocardium and CNR of the ventricular blood and myocardium were calculated from a quantitative region of interest analysis of these data. Additionally at 3.0 T, long axis and 4-chamber cine as well as "dark blood" imaging are demonstrated with sequence and parameter settings comparable to current state of the art for cardiac evaluation at 1.5 T. Results: The 3.0 T data consistently demonstrates increases in SNR when all imaging conditions are closely matched but the increase has a large variability ranging from 20 to 85% depending on the radiofrequency coil configuration. Ventricular blood-myocardium CNR greater than 30 is obtained at 3.0 T, which is comparable to an optimized 1.5 T acquisition despite the specific absorption rate limitation of flip angle to nearly one half the value. The increased SNR at 3.0 T improves detection of fine anatomic detail, such as the chordae tendineae and mitral valve structure. Conclusions: Increased specific absorption rate can be a limiting fact; however, we have demonstrated that 3.0 T cardiac imaging shows gains in SNR while maintaining the CNR. The SNR gain is advantageous, and phased array coil technology is key for improving cardiac magnetic resonance imaging at 3.0 T.