Characterization of cerebral artery flow waveforms in healthy subjects can help establish a comparison baseline in studies of cerebrovascular disease, and provide key inputs for computational fluid dynamics (CFD) model studies. Previous studies have primarily focused on characterizing arterial pulsatility in the carotid arteries [1-3]. Compliance-induced changes in pulsatility features along the internal carotid artery (ICA) and extending to the middle cerebral artery (MCA) remain unclear. The purpose of this study was to characterize cerebral arterial pulsatility and volumetric flow waveforms along both the ICA and MCA in adult volunteers using 4D flow MR imaging. 4D flow MRI with volumetric coverage (figure 1a) of the major intracranial vessels from proximal ICA to distal MCA was performed on a clinical 3T MRI scanner (Siemens, Erlangen, Germany) on 22 healthy volunteers (12 female, mean age 39.3±15.3 years, 19-60 years). Subjects were further divided into a young subgroup with age < 40 years (n=11, mean age 25.5±4.6 years) and an older subgroup with age > 40 years (n=11, mean age 54.1±6.3 years). 4D flow MRI sequence parameters were as follows: TR/TE= 5.2/2.8ms, flip angle= 15°, velocity sensitivity= 80 cm/s, voxel size = (1.1-1.4) mm3, temporal resolution= 41.6 ms, acquisition time= 15-20 minutes depending on the heart rate. 4D flow MRI data were preprocessed using a customized program (Matlab, The Mathworks, MA, USA) as described previously [4]. For each subject, fourteen 2D analysis planes (figure 1b, seven on each side, ICA: C2, C3, C4 and C7; MCA: M1-1, M1-2 and M1-3 at the location of proximal, middle and distal M1 segment) were extracted from a 3D phase-contrast MR angiogram and then were analyzed using an interactive flow analysis tool which allows frame-by-frame updating of the flow contours. Flow waveforms were derived from the 2D analysis planes and used to calculate pulsatility index (PI) from the following equation [1],$$PI=\frac{Fmax-Fmin}{Fmean},$$ where Fmax, Fmin and Fmean are peak systolic, minimum diastolic and mean flow rate, respectively. Volumetric flow rate was interpolated using spline interpolation from 4D flow MRI data, averaged over the left and right plane for all subjects (figure 2), and pulsatility indices were extrapolated from these waveforms (figure 3). Older adults showed decreased overall flow in comparison with the younger age group. For both subject groups, pulsatility index did not change significantly from proximal ICA to distal MCA. However, we observed a noticeable decreased pulsatility from proximal to distal carotid siphon (ICA C4-C7) in the young adult subgroup (figure 3a). In addition, older adults showed a trend towards decreased overall pulsatility in comparison with the younger age group possibly due to stiffening of vessels over time (0.61±0.28 versus 0.82±0.23, p = 0.08) . Standard deviation error bars indicate wide variability between subjects, and widening variability throughout the smaller MCA vessels suggests limits in MRI resolution as vessel size decreases. Flow waveforms and pulsatility index values were successfully extracted from 4D flow MRI data to provide a quantitative characterization of cerebral arterial waveforms in healthy adult subjects. Our findings demonstrated no significant changes of cerebral arterial pulsatility from proximal ICA to distal MCA, although a decreasing trend was observed from ICA C4 to C7 in young adults. As blood traverses the cerebral vasculature, the vessel lumen narrows as it reaches the MCA. Narrowed lumen and overall lower vessel velocities cause decreases in magnitude intensity, which correspondingly limit the accuracy of the user-dependent definition of the vessel lumen. Nevertheless, preliminary MRI simulations using cerebrovascular CFD models suggest that pulsatility index is less sensitive to spatial resolution compared with absolute volume flow rates. Further studies are warranted to better characterize cerebral flow dynamics with larger cohorts and higher imaging resolution.