Cancer imaging and treatment monitoring with color magnetic particle imaging

Abstract

Magnetic particle imaging (MPI) is emerging as a highly promising non-invasive tomographic imaging modality for cancer research. Superparamagnetic iron oxide nanoparticles (SPIONs) are used as imaging tracers in MPI. By exploiting the relaxation behavior of SPIONs, the capabilities of MPI can also be broadened to functional imaging applications that can distinguish different nanoparticles and/or environments. One of the important applications of functional MPI is viscosity mapping, since certain cancer types are shown to have increased cellular viscosity levels. MPI can potentially detect these cancerous tissues through estimating the viscosity levels of the tissue environment. Another important application area of MPI is temperature mapping, since SPIONs are also utilized in magnetic fluid hyperthermia (MFH) treatments and MPI enables localized application of MFH. To achieve accurate temperature estimations, however, one must also take into account the confounding effects of viscosity and temperature on the MPI signal. This dissertation studies relaxation-based viscosity and temperature mapping with MPI, covering the biologically relevant viscosity range (<5 mPa·s) and the therapeutically applicable temperature range (25-45!C). The characterization of the SPION relaxation response was performed on an in-house arbitrarywaveform magnetic particle spectrometer (MPS) setup, and the imaging experiments were performed on an in-house MPI scanner. Both the MPS setup and the MPI scanner were designed and developed as parts of this thesis. The effects of viscosity and temperature on relaxation time constant estimations were investigated, and the sensitivities of MPI to these functional parameters were determined at a wide range of operating points. The relaxation time constants, t’s, were estimated with a technique called TAURUS (TAU, t, estimation via Recovery of Underlying mirror Symmetry), which is based on a linear relaxation equation. Although the nonlinear relaxation behaviors of the SPIONs are highly dependent on the excitation field parameters, SPION type, and the hardware configuration, the results suggest that one-to-one relation between the estimated t and the targeted functional parameters (i.e., viscosity or temperature) can be obtained. According to these results, MPI can successfully map viscosity and temperature, with higher than 30%/mPa/s sensitivity for viscosity mapping and approximately 10%/!C sensitivity for temperature mapping, at 10 kHz drive field frequency. In addition, the results suggest that the simultaneous mapping of viscosity and temperature can be achieved by performing multiple measurements at different drive field frequencies and/or amplitudes. Overall, these findings show that hybrid MPI-MFH systems offer a promising approach for real-time monitored and localized thermal ablation treatment of cancer. The viscosity and temperature mapping capabilities of MPI via relaxation time constant estimation can provide feedback for high accuracy thermal dose adjustment to the cancerous tissues, thereby, increasing the efficacy of the treatment.