Adaptive vessel tracking: Automated computation of vessel trajectories for improved efficiency in 2D coronary MR angiography

A new method was investigated for improving the efficiency of ECG-gated coronary magnetic resonance angiography (CMRA) by accurate, automated tracking of the vessel motion over the cardiac cycle. Vessel tracking was implemented on a spiral gradient-echo pulse sequence with sub-millimeter in-plane spatial resolution as well as high image signal to noise ratio. Breath hold 2D CMRA was performed in 18 healthy adult subjects (mean age 46 +/- 14 years). Imaging efficiency, defined as the percentage of the slices where more than 30 mm of the vessel is visualized, was computed in multi-slice spiral scans with and without vessel tracking. There was a significant improvement in the efficiency of the vessel tracking sequence compared to the multi-slice sequence (56% vs. 32%, P < 0.001). The imaging efficiency increased further when the true motion of the coronary arteries (determined using a cross correlation algorithm) was used for vessel tracking as opposed to a linear model for motion (71% vs. 57%, P < 0.05). The motion of the coronary arteries was generally found to be linear during the systolic phase and nonlinear during the diastolic phase. The use of subject-tailored, automated tracking of vessel positions resulted in improved efficiency of coronary artery illustration on breath held 2D CMRA.