Abstract We developed a physical phantom that mimics laminar biological microstructure with controlled geometric and diffusional properties. The primary goal of this study is to characterize the diffusion properties of the phantom and to assess how laminar microstructural geometry influences diffusion-weighted (DW) magnetic resonance imaging (MRI) signals. The phantom consists of stacked polyethylene foils forming water layers with thicknesses ranging from ~ 1 to 16 μm. Water diffusion within these layers is free in-plane but restricted perpendicular to the boundaries. We initially modeled the perpendicular component of the signal using a Laplacian spectral approach, while the parallel component was described by standard Gaussian attenuation. Although this model captures the overall signal behavior, the measured data reveal systematic deviations attributable to local structural inhomogeneities, such as the formation of water pockets between foil layers. To account for these effects, we introduce an extension of the model that incorporates spatial variations in layer spacing. Our data, acquired on a 7 T MRI scanner, show good agreement with the extended model and suggest the presence of such water pockets. Their planar dimensions, which range from 28 to 1500 μm, depend on foil compression and scale proportionally with the thickness of the water layer. These findings show that accurate diffusion models are needed to correctly estimate the phantom’s microstructure, and that the developed phantom is a useful tool for studying laminar diffusion. The remaining mismatch between the model and the data suggests that additional structural factors, not captured by the model, also influence the measurements.
Łabieniec et al. (2026) studied this question.