Quantitative imaging of magnetic nanoparticles by magnetorelaxometry with multiple excitation coils

Abstract

New therapies against cancer based on magnetic nanoparticles (MNPs) require a quantitative spatially resolved imaging of MNPs inside a body. In magnetorelaxometry (MRX), a distribution of nanoparticles can be quantified non-invasively by measuring its relaxation after removal of an external magnetizing field. Conventionally, in MRX the sample is exposed to a homogeneous magnetizing field resulting in a quantitative reconstruction with rather poor spatial resolution. Theoretical work suggests an improvement of spatial resolution may be achieved by a sequential application of inhomogeneous fields magnetizing only parts of a sample. Here, we experimentally demonstrate the feasibility of this approach by reconstructing a nanoparticle distribution inside a compact three-dimensional volume phantom made of 54 gypsum cubes (1 cm(3) cube(-1)), of which 12 gypsum cubes were filled with MNPs. Using 48 small excitation coils surrounding the phantom, a sequence of MRX signals was obtained where only those MNPs near an individual coil contribute. By combined evaluation of these 48 MRX measurements, the positions and content of the 12 MNP-filled cubes could be determined accurately with a deviation below 4%, while by conventional homogeneous MRX only the MNP content was reconstructable with a deviation of about 9%. The results demonstrate the improvement of quantitative MRX imaging by using sequential activation of multiple magnetizing fields.

Quantitative model selection for enhanced magnetic nanoparticle imaging in magnetorelaxometry

“…Because of the different orientations and the spatial arrangement of the sensors, we are able to measure the magnetic induction vector at different positions. 13,23 A sample which contains a certain spatial distribution of MNP is considered. We virtually tessellate this sample into V voxels.…”

“…In our simulations, we employ the distributed coil array MRX setup at PTB. 23 This setup has 48 (N = 48) planar coils with a diameter of about 1 cm. Twelve coils are positioned between the SQUIDs and the sample with the unknown MNP distribution, 12 coils are positioned below the sample, and 2×12 coils are positioned perpendicular to the other coils, with the sample between them [see Fig.…”

“…20,21 The magnetizing field is generally produced by a Helmholtz coil 22 or a distributed coil array. 23 The inverse problem requires a forward model to generate simulated measurements. These are then compared to the MRX measurements in the inverse problem.…”

Quantitative and binding-specific imaging of magnetic nanoparticle distributions

“…Thus, it contains information about the binding state of the MNPs. Recently, we experimentally demonstrated quantitative imaging of an MNP distribution by sequential MRX measurements using spatially varying magnetization field patterns [1]. Here, we extend this approach by a phenomenological data model [2] in order to identify different binding states of the MNPs in the images of the MNP distribution.…”

Extended arrays for nonlinear susceptibility magnitude imaging

“…Over the past decade, researchers have begun to understand how to exploit the unique magnetic properties of mNPs for medical imaging applications [21, 22]. The three primary methods that have emerged are magnetic susceptibility imaging [3–5, 14, 20, 30], magnetic relaxometry (MRX) [1, 6, 15, 17, 18, 24–26, 28, 31], and magnetic nanoparticle imaging (MPI) [2, 10–12, 16, 23, 29, 27]. Our group has recently introduced several new algorithms for mNP imaging in methods we call susceptibility magnitude imaging (SMI) [9], spectroscopic AC susceptibility imaging (sASI) [8], and nonlinear susceptibility magnitude imaging (nSMI) [7].…”