Monte Carlo simulations of transverse relaxation for characterization of physicochemical properties of superparamagnetic iron oxide nanoparticles

Abstract

Superparamagnetic iron oxide nanoparticles are used in MRI as T2 or T2* contrast agents and, in combination with relaxometry, enable quantitative analysis of physiological and pathological processes. However, the induced changes in relaxation times are influenced by a complex interplay of the contrast agents’ physicochemical properties. Here, an open-source Monte Carlo simulation pipeline was implemented, which enables the char­ acterization of these relaxation time changes caused by MRI contrast agents. The simulation tool was validated by showing that simulated relaxation times for iron oxide particles matched the solutions of analytical models of the respective diffusion regimes. For comparison with relaxometry measurements, T2 and T2* of four MRI contrast agents Ferucarbotran, FeraSpin XL, magnetite nanohexagons (MNH@OH) and magnetite nanocubes (MNC@OH) were simulated, using three approaches for modeling contrast agent size and composition: 1) uniform particle sizes using the median hydrodynamic radii; 2) distributed radii corresponding to measured hydrodynamic radius distributions; 3) size-distributed magnetite cores in a coating layer of uniform radius. The simulation accurately reproduced measured relaxation times when appropriate modeling strategies for contrast agent size and composition were used. For FeraSpin XL and MNH@OH, using uniform radii provided good estimates of relaxation times, which was further improved by using the size distributions. For MNC@OH, discrepancies in simulated and measured T2 for all approaches were attributed to particle aggregation. For Ferucarbotran, coating and size distribution of the core had to be considered to match experimental data.