Theoretical considerations on the quantification of iron oxide labeled cells in vivo

Introduction Iron oxide contrast-enhanced MRI has become a commonly used tool in molecular and cellular imaging. By labeling cells with iron oxide nanoparticles, it is possible to create a strong local magnetic field inhomogeneity. This leads to signal attenuations in T2 and T2* weighted MR images on the site of the cell. These hypointense spots can be utilized on longitudinal studies to sensitively track labeled cells, providing a powerful non-invasive tool for biological questions. Besides other problems such as ambiguity of the signal attenuation, another major issue which is not successfully solved until now is the quantification of these labeled cells. Due to the insufficient resolution of MRI, only strongly labeled and sparsely distributed cells can be detected separately. This allows quantification of cell numbers by direct counting. For denser packing and lower cell labels, the measurement of T2 and T2* is suggested for quantification [1-4]. All of these studies, however, suffer from the unknown amount of iron oxide internalized by the cells during in vivo situations. To circumvent this problem two approaches are used. Either, the total iron amount per volume is quantified [1, 2, 4] or calibration curves are used to determine the cell density from the relaxation times [3]. These calibration curves must be determined by in vitro experiments or histological studies. One possible solution to avoid calibration curves is the measurement of both relaxation times. Another one is measuring the dependence of the T2 relaxation time on the inter-echo time of a multi spin echo sequence. In this work we investigate these possibilities by studying theoretical models for the transverse relaxation in the presence of diffusion on the typical parameter range of in vivo situations.