Automatic segmentation of myocardium at risk in T2-weighted cardiovascular magnetic resonance

Abstract

SummaryThis study presents a newly developed automatic method for segmentation of myocardium at risk using a priori knowledge on perfusion territories. The new automatic method shows a low bias and high correlation to manual delineation and a bias and correlation closer to inter observer variability for manual delineation than three existing threshold methods, 2SD from remote, FWHM and Otsu.

Association between epicardial adipose tissue volume and myocardial salvage in patients with a first ST-segment elevation myocardial infarction: An epicardial adipose tissue paradox

“…The infarct size was defined as the extent of Core. On T2-weighted images, myocardial tissue with an SI of at least 2 SD above the mean SI obtained in the remote non-infarcted myocardium was considered the area at risk (AAR) [15]. MS and MSI were then estimated by the following formulae [16]: MS = AAR minus Core MSI = MS divided by AAR Intramyocardial hemorrhage was defined as a dark area within the hyperenhanced area on T2-weighted images that was considered to belong to the AAR.…”