Temporal covariance analysis of first-pass contrast-enhanced myocardial magnetic resonance images.

In this paper a temporal covariance method designed to analyze a Magnetic resonance (MR) image sequence of myocardial perfusion is presented. This method is used to map the first-pass transit of a contrast agent (Gd-chelates) through the heart. A map of bolus transit delay is constructed pixel by pixel corresponding to a myocardial reference using a temporal covariance measure. The resulting covariance map is a parametric image representing regions with different temporal dynamics. The proposed method is evaluated in 14 patients with coronary artery disease and eight healthy volunteers. Under rest and stress, covariance method is able to reveal a perfusion defect in stenosed coronary-artery-related myocardium. Furthermore, the method presents the advantage of its easy implementation and real-time parametric map construction.