Accurate measurement of pulsatile flow velocity in a small tube phantom: comparison of phase-contrast cine magnetic resonance imaging and intraluminal Doppler guidewire.

PURPOSE: We compared the accuracy of magnetic resonance imaging (MRI) measurements of pulsatile flow velocity in a small tube phantom using different spatial factors versus those obtained by intraluminal Doppler guidewire examination (as reference). MATERIALS AND METHODS: We generated pulsatile flow velocities averaging about 20-290 cm/sec in a tube of 4 mm diameter; we performed phase-contrast cine MRI on pixels measuring 1.00(2)-2.50(2) mm(2). We quantified spatial peak flow velocities of a single pixel and a cluster of five pixels and spatial mean velocities within regions of interest enclosing the entire lumen in the phantom's cross-section. Finally, we compared the measurements of temporally mean and maximum flow velocity with the Doppler measurements. RESULTS: Linear correlation was excellent between both measurements of spatial peak flow velocities in one pixel. The highest spatial resolution using spatial peak flow velocities of a single pixel allowed the most accurate MRI measurements of both temporally mean and maximum pulsatile flow velocity (r = 0.97 and 0.99, respectively: MRI measurement = 0.95x + 8.9 and 0.88x + 24.0 cm/s, respectively). Otherwise, MRI measurements were significantly underestimated at lower spatial resolutions. CONCLUSION: High spatial resolution allowed accurate MRI measurement of temporally mean and maximum pulsatile flow velocity at spatial peak velocities of one pixel.

Spatial factors for quantifying constant flow velocity in a small tube phantom: comparison of phase-contrast cine-magnetic resonance imaging and the intraluminal Doppler guidewire method.

PURPOSE: We examined the spatial factors influencing magnetic resonance (MR) flow velocity measurements in a small tube phantom and used the same measurements obtained with an intraluminal Doppler guidewire as reference. MATERIALS AND METHODS: We generated constant flow velocities from approximately 40 to 370 cm/s in a tube 4 mm in diameter. We then performed segmented k-space, phase-contrast cine-MR imaging to quantify spatial peak flow velocities of one pixel and of five adjacent pixels as well as spatial mean velocities within regions of interest in a cross section of the phantom. Pixel dimensions ranged from 1.00 x 1.00 mm to 2.50 x 2.50 mm. We compared the MR measurements with the temporally averaged Doppler spectral peak velocities. RESULTS: For one pixel (r > 0.99: MR flow velocity for pixel dimension 1.00 x 1.00 mm = 1.03x + 9.8 cm/s), the linear correlation was excellent between flow velocities by MR and Doppler guidewire methods. However, for the five adjacent pixels, MR measurements were significantly underestimated using pixels 1.25 x 1.25 mm to 2.50 x 2.50 mm and for mean velocities for all pixel dimensions. CONCLUSION: Relatively high spatial resolution allows accurate MR measurement of constant flow velocity in a small tube at spatial peak velocities for one pixel.