Coronary flow and flow reserve in canines using MR phase difference and complex difference processing.

Abstract

Coronary artery disease continues to be the leading cause of death for adults in the United States. Magnetic resonance imaging (MR) has the potential to dramatically impact the diagnosis of heart disease by noninvasively providing a wide range of anatomic and physiologic information. Previous research has shown that coronary flow, one component of a complete examination, can be accurately measured in the left anterior descending artery in vivo. The current work validates MR flow measurements in canine circumflex arteries using transit time ultrasound as a standard. The circumflex artery experiences greater in-plane motion and is a more stringent test for flow measurement accuracy. This work also compares two methods of processing MR velocity data, phase difference and complex difference techniques, and examines the sources of error present in the animal validation model. Phase difference processing with a 30% magnitude threshold best matched the mean ultrasound flow values (30% PD = 1.04 x US + 1.49, r = 0.94), but it was very sensitive to vessel boundary identification. The complex difference process was less sensitive to vessel boundary identification and correlated well with the transit time ultrasound despite systematic underestimations. The reasons for the discrepancies are shown to stem from a number of possible sources including variability of the ultrasound standard, low signal-to-noise ratios in the MR images, sensitivity of the MR technique to vessel boundary identification, and motion artifacts in the images.