Measurements of a perivascular spaces magnetic resonance imaging physical phantom and correspondent digital reference object model

Magnetic Resonance Imaging (MRI) visible perivascular spaces (PVS) have been associated with age, decline in cognitive abilities, interrupted sleep, and markers of small vessel disease. Therefore, several computational methods have been developed for their assessment from brain MRI. But the limits of validity of these methods under various spatial resolutions, and the accuracy in detecting and measuring the dimensions of these structures have not been established. We developed a digital object identifier (DRO) and validated it using a physical phantom from UMC Utrecht to answer these questions. Our in-silico phantom or PVS-DRO consist on cylinders with varying diameters from 0.2 to 3mm each one confined to a space of 15x15x15 mm as per “enlarged MRI-visible PVS” are defined in the literature. This dataset contains the source code to generate such in-silico cylinders, an excel spreadsheet with the measurements of the cylinders of the UMC Utrecht physical phantom from thresholding the output of two vesselness filters: Frangi and RORPO, and the source code to generate such measurements. The excel file also contains Bland-Altman analyses to inform on the accuracy and behaviour of the filters in scenarios of varying spatial resolution and cylinders’ orientation. The file relates to the upcoming publication Duarte Coello, R., Valdés Hernández, M.C., Zwanenburg, J.J.M., van der Velden, M., Kuijf, H.J., De Luca, A., Bernal Moyano, J., Ballerini, L., Chappell, F.M., Brown, R., Biessels, G.J., Wardlaw, J.M. (in submission). "On the detectability and accuracy of computational measurements of enlarged perivascular spaces from magnetic resonance images".