1045Blood/myocardium contrast-to-noise ratio for cardiac morphology and function imaging increases with field strength

Abstract

Objective: To investigate signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) between blood and myocardium in fast gradient echo (FGRE) and steady state free precession (SSFP) sequences for magnetic field strengths of 1.5T, 3T, and 7T and to derive adjusted flip angles. Methods: Signal calculations for FGRE and SSFP were done in Matlab (The Mathworks, Natick, MA) using common signal equations and Bloch simulations. Relative spin densities were set to 0.7 and 0.95 for muscle and blood [2]. For the simulations, T1 and T2 values of myocardium and blood were derived from the literature for field strengths of 1.5T, 3T, and 7T [3-10]. Mean T1 for myocardium/blood of 992ms/1540ms (1.5T), 1250ms/1741ms (3T), and 1938ms/2277ms (7T) were used. Mean T2 for myocardium/blood were 52ms/290ms (1.5T), 47ms/275ms (3T), and 37ms/242ms (7T). TR/TE of 5ms/2.5ms were applied. T2* effects were neglected due to the use of short echo times. CNR(blood/myocardium) was calculated as SNR(blood)-SNR(myocardium). Signal differences between blood and myocardium were calculated as (SI(blood)-SI(myocardium)). Results: Figure 1 shows blood/myocardium CNR values for FGRE assuming a) ideal slice profiles and b) Bloch simulated and imperfect slice profiles. CNR adjusted flip angles were a) 3°/3°/3° and b) 6°/5°/5° for FGRE at 1.5T/3T/7T. Figure 2 shows Bloch simulated blood/myocardium CNR for SSFP at presuming a) ideal and b) imperfect slice profiles. CNR-adjusted flip angles were a) 56°/51°/42° and b) 84°/78°/66° for 1.5T/3T/7T. Figure 3 shows signal differences for blood and myocardium assuming imperfect slice profiles. While FGRE benefits from the higher field strength, SSFP blood myocardium signal difference decreases. Simulated blood/myocardium CNR over different flip angles at 1.5T/3T/7T for cardiac FGRE assuming a) ideal slice profile and b) real slice profile employing bandwidth cropped sinc-RF-pulses. Simulated blood/myocardium CNR over different flip angles at 1.5T/3T/7T for cardiac SSFP assuming a) ideal slice profile and b) real slice profile employing bandwidth cropped sinc-RF-pulses. Simulated blood/myocardium signal difference for a) FGRE and b) SSFP at 1.5T/3T/7T using different flip angles and assuming imperfect slice profiles.