MRI phase contrast velocity and flow errors in turbulent stenotic jets

PURPOSE: To clarify the use of MRI phase contrast (PC), as an alternative to Doppler echocardiography, when measuring high-velocity turbulent jets associated with stenotic valvular disease. MATERIALS AND METHODS: In vivo PC aortic stroke volume (SV) was compared with ventricular SV in 31 patients with moderate to severe aortic stenosis (AS). Two in vitro pipe experiments were conducted to evaluate errors in steady stenotic and nonstenotic turbulent flows. RESULTS: The average in vivo error in SV was -24% in the left-ventricular (LV) outflow tract (LVOT) and -41% in the aortic root. Errors were most prominent in patients with the highest Doppler peak velocities. In vitro nonstenotic flow experiments showed accurate flow measurement with an average error of 1.8%. Significant errors were found in the in vitro stenotic flow, which reduced with shorter echo times (TE): average error -166/-67/-25/-13/-8.8% for TEs of 4.8/4.0/3.3/2.2/2.0 msec. In both the in vivo and in vitro stenotic experiments the errors were associated with signal loss in the flow-compensated magnitude image. CONCLUSION: Signal loss is associated with flow errors in stenotic jets. Current clinically available PC pulse sequences with TE >2 msec may not accurately quantify flow for severe lesions.