Magnetic particle imaging (MPI) generates signal from the non-linear magnetic response of superparamagnetic iron oxide (SPIO) nanoparticles. This study compared the performance of ten commercially available iron oxide tracers across a range of magnetic core sizes, hydrodynamic diameters, and surface coatings. Tracer performance was evaluated using magnetic particle relaxometry (MPR) peak signal intensity and two image-based 2D MPI metrics: total signal and maximum signal intensity per microgram of iron. The authors report that Synomag-D tracers yielded the highest MPR peak signal and highest maximum MPI signal per µg iron, while ProMag, a micron-sized iron oxide particle (MPIO), produced the highest total MPI signal per µg iron despite a low MPR peak signal.
Ten commercially available tracers were included to span a broad range of physicochemical characteristics: magnetic core size, hydrodynamic diameter, and surface coating. The manuscript lists these broad categories of variation but does not provide detailed composition, manufacturer catalog identifiers, or the full quantitative values for each tracer in the abstract. Those detailed tracer-level properties and experimental concentrations are not reported in the abstract and would need to be consulted in the full text for experimental replication.
Two complementary approaches were used to evaluate tracer performance. Magnetic particle relaxometry (MPR) measures the peak signal intensity from the tracer's magnetic response under relaxometry conditions. Image-based evaluation used two 2D MPI-derived metrics calculated from images: the total MPI signal (integrated signal across the image region of interest) and the maximum MPI signal intensity (single-voxel or pixel peak within the image), each normalized to micrograms of iron. The abstract does not report the specific instrument models, imaging parameters, or image processing workflows; those methodological details are not provided in the abstract and must be sought in the full manuscript.
Among the tracers tested, Synomag-D tracers achieved the highest MPR peak signal intensities and the highest maximum signal per µg iron in MPI images. In contrast, ProMag, categorized as an MPIO, showed a low MPR peak signal intensity but produced the highest total MPI signal per µg iron. These findings indicate that different tracers can rank differently depending on the metric used: relaxometry peak versus integrated image signal.
The study assessed how well MPR peak signal intensity predicts image-based performance. A strong correlation was observed between MPR peak signal and MPI maximum signal intensity (R2 = 0.94), indicating that MPR peak is a good predictor of the peak image intensity across the tracers tested. By contrast, the correlation between MPR peak signal and MPI total signal intensity was weak when all tracers were included (R2 = 0.38). However, removing micron-sized iron oxide particles (MPIOs) from the analysis markedly improved this correlation (R2 = 0.91). This pattern suggests that MPIOs can produce disproportionately large total image signal relative to their MPR peak signal, and that excluding such outliers makes MPR a more reliable predictor of total image signal among SPIO tracers.
The authors interpret the data to mean that magnetic particle relaxometry generally predicts tracer performance, particularly for predicting maximum image intensity, but it does not fully reproduce the complex conditions encountered during imaging. Differences in how tracers contribute to integrated image signal—especially for micron-scale iron oxide particles—highlight that relaxometry alone may be insufficient when assessing tracer suitability for specific MPI applications. The study therefore supports a combined evaluation strategy using both MPR and image-based metrics to capture complementary aspects of tracer behavior.
The main conclusion is that comprehensive tracer evaluation for MPI requires combining relaxometry and image-based metrics. MPR is a strong predictor of maximum MPI image intensity (high R2), but image-based measures, particularly total signal, can diverge from relaxometry predictions when particle types such as MPIOs are present. The authors recommend integrating both measurement approaches to better assess tracer performance, acknowledging that relaxometry does not completely replicate imaging conditions.
The authors declared no competing interests. Funding was reported from the Canadian Institutes of Health Research and the Natural Sciences and Engineering Research Council. This work is presented as a preprint and has not been peer reviewed; readers should interpret results accordingly.
The abstract provides high-level findings and key correlations but omits several experimental specifics in this summary, including exact tracer identities and concentrations, instrumentation and imaging parameters, relaxometry acquisition conditions, statistical methods beyond the reported R2 values, and quantitative tables or figures. Those methodological and quantitative details are not reported in the abstract and should be consulted in the full preprint PDF for replication or deeper technical interpretation.