Imaging Characterization of MPI Tracers Employing Offset Measurements in a 2D Magnetic Particle Spectrometer

We developed a method to characterize the imaging performance of MPI tracers from virtual MPI measurements that can be synthesized using measurement data from a Magnetic Particle Spectrometer (MPS) at different static magnetic field offsets. MPI system functions were obtained from measurements on a FeraSpin TM R (nanoPET GmbH, Berlin) sample in a 2D MPS comprising two excitation coils and two receive coils. Software phantoms of spatial MPI tracer distributions with different sizes and shapes were constructed. With the measured MPI system function, a synthetic MPI measurement of the software phantoms was simulated. By adding noise to the virtual MPI data, the detection limit of each harmonic in dependence of the noise level was obtained. An MPI reconstruction of the virtual tracer distribution was performed using the virtual MPI data as input. With this method we found the highest tolerable noise level at which it was still possible to distinguish the objects in the phantom. These findings were compared to predictions based on the frequency components that were used in the reconstruction process. This method provides a valuable link between pure spectroscopic characterization and time consuming MPI phantom experiments.