Assessing cerebral arterial pulse wave velocity using 4D flow MRI

Figure 1.. Schematic overview of the process to obtain PWV. The complex difference (CD) image is used to generate a centerline representation of the vascular tree. A user-defined selection determines the centerlines of interest (CL) in which velocity curves, calculated from velocity images, are automatically extracted. The velocity curves are stored in a two-dimensional matrix. A distance calculation is also performed on the CL volume, generating a unique distance (r) for each cross-section that is stored as vector with the same cross-section ordering as the velocity matrix. A model is fitted to the velocity matrix and the distance vector to obtain the PWV as well as the velocity waveform that best explain the observed data.

Figure 2.. (a) Example angiographic complex difference (CD) image; (b) the raw centerline structure; (c) the corresponding segmented centerline structure (CL) with calculated vascular depth; (d) the average velocity waveforms obtained in the most proximal (depth <50 mm) as well as in the most distal aspects (depth >250 mm) of the visible cerebral vasculature; (e) the initial guess and final estimation of the arterial velocity waveform obtained in the optimization procedure.

Figure 3.. (a) Histogram illustrating the average number of cross-sections available as a function of vascular depth; (b) individual PWV values for young and old participants. Among the elderly, individuals without previous vascular disease, dementia or diabetes (N = 21) are marked in blue; (c) stability analysis performed by calculating PWV on half the number of cross-sections; (d) stability analysis performed by calculating PWV using only main routes of vascular supply. The red line in (c) and (d) illustrates perfect agreement.