Increasing reliability of cranial blood flow velocity measurements in preterm infants

Background Long-term prognosis of preterm infants is strongly related to the occurrence of brain injury, in which inadequate perfusion seems to play a key role. Using high-end ultrasound machines with high-frequency probes visualization of microvessels and regional variation is possible nowadays. The next step in neonatal neuro-imaging would be quantification of ultrasound parameters. Objective To assess whether flow velocity measurements of preterm cerebral perfusion obtained by Doppler ultrasound are accurate, we conducted two in vitro experiments using microvessel flow phantoms as a model for a preterm brain. Design/Methods Two in-house developed flow phantoms containing microvessels (inner diameter 160 and 700 microns) with attached syringe pumps, filled with blood-mimicking fluid, were used to generate non-pulsatile perfusion with peak flow velocities of 1-10 cm/s. Using presets developed for daily clinical practice, velocity measurements were performed using an Esaote scanner (MyLab 70 XVG, Genova, Italy). In the second part of the study the linear ultrasound probe was connected to an other Esaote ultrasound machine (MyLab 60, Genova, Italy; Doppler settings identical to MyLab 70 XVG) and calibrated. Results Microvessels with velocities as low as 1cm/sec were adequately visualized with a linear ultrasound probe. With a convex probe velocities <2 cm/sec could not be depicted. Intra-observer reproducibility for velocity and diameter measurements was good (ICC 0.997 and 0.914). However, velocity was overestimated up to 3-fold. Significant differences were seen in velocity measurements when using steer angle correction and in vessel diameter estimation (p<0.05). In the second part of the study, calibration of the linear ultrasound probe resulted in highly reproducible and reliable maximum velocity results of non-pulsatile flow. Increasing Doppler frequency provides more reliable results (figure 1). No significant differences were seen between high versus low pulse repetition frequency. Applying steer angle correction gives most reliable results (p<0.05). Conclusion. Visualization of microvessel size catheters mimicking small brain vessels as small as 160 micrometers is feasible. Reproducible velocity results can be obtained, although important overestimation of the values is observed if standard presets are used. Calibration leads to highly reproducible results. Before velocity estimates in small brain vessels can be used in clinical practice, calibration of the ultrasound machine for any specific Doppler purpose using a flow phantom is highly recommended.