The Effect of Different Gadolinium Delivery Rates on Objective and Subjective Image Quality in Cardiac Perfusion Magnetic Resonance Imaging.

Abstract

Background: Stress perfusion cardiac magnetic resonance imaging (MRI) enables accurate detection of myocardial ischemia. While contrast injection parameters, particularly iodine delivery rate, are known to influence image quality in CT, their impact in cardiac MRI remains insufficiently studied. This study aimed to evaluate the effect of different gadolinium-based contrast agent (GBCA) injection rates on objective and subjective image quality in perfusion cardiac MRI in patients with suspected ischemic heart disease.

Methods: Ninety patients scheduled for a stress perfusion cardiac MRI were retrospectively included and allocated to 3 groups based on the GBCA injection rate used (eg, group 1: 6 mL/s, group 2: 4 mL/s, and group 3: 2 mL/s). Blood-pool signal intensity was extracted on a frame-by-frame basis to generate signal-intensity time curves. Signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) were calculated for all relevant regions, including SNRischemia, SNRblood, SNRmyocardium, CNRischemia-blood, CNRmyocardium-blood, and CNRischemia-myocardium. Subjective image quality was scored on a 5-point Likert scale, blindly assessed in consensus by 2 cardiovascular radiologists.

Results: Blood-pool peak intensity was the highest in the group with flow rates of 6 mL/s. Peak signal intensity was for the 6, 4, and 2 mL/s group, respectively, 17525±3616, 15213±3416, and 15534±2990 a.u. (P=0.019) and post hoc comparison demonstrated a significant difference between the 6 mL/s and 4 mL/s group (P=0.027). No significant differences in signal intensity of the myocardium, blood pool and the infarcted region were found between groups (P=0.475, 0.399 and 0.519, respectively). Image noise was significantly higher in group 1 compared to group 3 (P value=0.036). SNR of myocardium, blood, and ischemia was not significantly different between groups (P=0.525, 0.595, and 0.453, respectively). CNR for ischemia-blood, myocardium-blood and ischemia-myocardium were the highest in group 3, but not significantly different (P=0.373, 0.896, and 0.346, respectively). Subjective image quality and reader confidence were not significantly different between groups (P values of 0.985 and 0.842). Although contrast quality demonstrated a statistically significant difference across groups (P=0.026), this finding was not considered clinically relevant.

Conclusion: Higher GBCA injection rates resulted in higher blood-pool peak enhancement, yet did not translate into significant differences in objective image and subjective image quality. Lower injection rates appear sufficient to achieve adequate myocardial perfusion imaging, suggesting a plateau in image quality in which increasing flow rates do not show added benefit.