UC Berkeley

Magnetic particle imaging (MPI) is a novel medical imaging modality that spatially detects a tracer of superparamagnetic iron oxide nanoparticles (SPIOs) with high sensitivity, contrast, and no tissue penetration limitations. MPI has great potential for safer angiography, in vivo cell tracking, and cancer detection, among other applications. Current MPI theoretical descriptions and reconstruction techniques make an adiabatic assumption that the SPIO tracer instantaneously follows the applied magnetic fields of the MPI scanner. This assumption is not strictly true, and we refer to SPIO magnetization delays as relaxation effects.

We begin by extending the x-space theory of MPI to include relaxation effects. We choose this MPI theory because it directly converts the temporal MPI signal to the image (spatial) domain, lending itself well to investigating how relaxation time delays translate into spatial effects. Using the non-adiabatic x-space theory and experimentally-measured data, we demonstrate that relaxation blurs the x-space image in the scanning direction.

Next, we study how we may design MPI scanning sequences to minimize relaxation-induced blurring. From the non-adiabatic x-space theory we derive a mathematical description of how this blur can vary with scanning parameters for a given relaxation time. We compare theoretical predictions to experimental data by measuring relaxation times and spatial resolution under various scanning conditions. Despite increased relaxation time delays with slower scanning conditions, we observe that relaxation-induced blurring can be minimized when scanning slower.

Finally, we derive a magnetic field-driven relaxation mechanism called magneto-viscous relaxation. This mechanism describes how the applied magnetic field creates a magnetic torque on the SPIO, inducing physical rotation of the SPIO to align with the field; however viscous resistance of the carrier liquid hinders this movement. We compare predicted relaxation times to measured values for a range of SPIO characteristics and scanning conditions.

In this dissertation, we show how relaxation can have deleterious effects on the MPI signal and image. We explore how relaxation-induced blurring and relaxation times may be minimized through improved SPIO characteristics and MPI scanning sequence design. In addition to improving MPI image quality, this important area of research can lead to future clinical applications. Using this knowledge and specially-designed MPI pulse sequences, we can exploit variations in relaxation behavior as a source of contrast, which will increase the diagnostic potential of MPI.