Calibration of diffusion MRI measurements using aqueous glycerin phantoms with controlled viscosity.

Abstract

To characterize the relationship between the apparent diffusion coefficient (ADC) and dynamic viscosity in aqueous glycerin phantoms across different concentrations and temperatures, and to establish a controlled calibration framework for diffusion MRI measurements. Eight glycerin-water solutions (10-60 wt%) were prepared, and viscosities were measured with capillary viscometers at 24, 37, and 40 °C. Diffusion-weighted imaging (DWI) was performed on a 3.0-T scanner using 15 b-values (0-1500 s/mm²). Voxel-wise ADC maps were calculated by mono-exponential fitting, and ADC-viscosity relationships were modeled using linear and inverse fitting approaches. Viscosity increased with concentration and decreased with temperature, consistent with the known physical behavior of glycerin-water mixtures. ADC values were inversely related to viscosity and showed strong relationships at all temperatures (R² > 0.98). High-concentration samples exhibited reduced signal-to-noise ratio and greater fitting instability, limiting reliable analysis in the highest-viscosity range. Aqueous glycerin solutions provide practical phantom materials for evaluating ADC-viscosity relationships under controlled conditions. By anchoring MRI-derived ADC measurements to standard viscometer data, this study established a reproducible calibration framework for diffusion MRI system evaluation across a defined viscosity range.