Comparison of Cerebral Vessel Pulsatility Using Vessel Diameter Algorithms

Neurodegenerative diseases such as Alzheimer's and neurological diseases such as stroke are linked to abnormal cerebrospinal fluid (CSF) flow. Normal CSF flow plays a role in clearing metabolic waste and is driven, in part, by the expansion and contraction of arteries adjacent to CSF-filled spaces. This study aims to measure and quantify arterial wall motion to better understand CSF dynamics and associated metabolic waste removal. We simulated microscopic data to measure vessel pulsations by creating movies of pulsing vessels with varied artifacts including translation and adjusting the noise-to-signal ratio. The average diameter, root mean square diameter, and pulsatility (quantified as the diameter interquartile range) error were calculated with three MATLAB algorithms and then compared against the ground truth. Among the algorithms, typically, the image intensity-based diameter measurement algorithm resulted in the highest pulsatility error, followed by the Radon transform-based algorithm, and finally the edge-detection-based algorithm. The comparison of these vessel diameter measurement algorithms highlights their respective limitations, offering insight into the most suitable algorithm for different datasets.