Late gadolinium-enhanced cardiovascular MRI at end-systole: feasibility study.

Abstract

The purpose of this article is to evaluate the image quality and infarct size of segmented late gadolinium-enhanced cardiovascular MRI at end-systole, compared with middiastole, in patients with sinus rhythm (SR) and to compare the image quality of end-systole images in patients with atrial fibrillation (AFib) to that of end-systole and middiastole images in patients with SR. Study patients (n = 121) were distributed according to heart rate and rhythm: SR with low heart rate (≤ 65 beats/minute), SR with intermediate heart rate (66-75 beats/minute), SR with high heart rate (≥ 76 beats/minute), and AFib. Image quality was graded on a 5-point scale, where 4 equals optimal and 0 equals not assessable. Global infarct size (percentage of left ventricle [LV] myocardium) in patients with SR with myocardial infarction was quantified using a visual quantitative approach with a 5-point scale and a semiautomatic method. End-systole imaging had higher image quality than did middiastole imaging for patients with SR with high heart rate, whereas middiastole imaging had higher image quality than did end-systole imaging for patients with SR with low heart rate (p < 0.05 for patients with SR with low heart rate, p = 0.60 for patients with SR with intermediate heart rate, and p = 0.001 for patients with SR with high heart rate). The quality of end-systole imaging in patients with AFib was not significantly different from that in patients with SR (p = 0.40 vs SR middiastole imaging and p = 0.38 vs SR end-systole imaging). The average difference of global infarct size was -0.3% and 0.2% of LV myocardium, and the limits of agreement were ± 2.4% and ± 3.3% of LV myocardium, for visual assessment and semiautomatic assessment, respectively. End-systole imaging can provide accurate diagnosis of myocardial infarction, comparable to middiastole imaging. The image quality of end-systole imaging is less susceptible to heart rate and rhythm compared with middiastole imaging.