Pitfalls in myocardial perfusion assessment with dynamic MR imaging after administration of a contrast material bolus in dogs

Rationale and Objectives. The authors evaluated the artifacts observed on myocardial perfusion curves derived from an inversion-prepared fast gradient-echo (GRE) imaging sequence in dogs after injection of a gadolinium-based contrast agent. Materials and Methods. Six mongrel dogs were divided into three groups. In groups 1 and 2, anesthesia was maintained with pentobarbital. Group 2 also received an intravenous injection of atropine (0.03 mg/kg). In group 3, anesthesia was maintained with isoflurane (1.0%). Imaging was performed on a 1.5-T magnetic resonance (MR) imaging unit (one section per heart beat, a 30 x 15-cm field of view, 10-mm section thickness, and 64-kHz bandwidth). Region-of-interest (ROI) markers were placed on the blood pool of the left intraventricular cavity, anterior wall of the left ventricle, and anterior to the chest wall to track respiratory motion. Results. In group 1, the signal intensity (SI) periodically increased during each inspiration due to respiratory sinus arrhythmia. The relation between the SI increase and the variation of the delay between images was demonstrated in vitro and by computer simulations. No periodic increase of the SI was observed when regular cardiac rhythm was maintained by pharmacologic inhibition of the vagal-mediated chronotropic response with either the addition of atropine to pentobarbital or the use of isoflurane as the anesthetic agent. Conclusion. In an inversion-prepared fast GRE sequence, respiratory sinus arrhythmia can induce periodic SI increase by varying the respiratory rate interval and delay between images.