Tailored regularization methods for multi-contrast magnetic particle imaging

Multi-contrast magnetic particle imaging (MPI) enables the determination of different contrasts in addition to the particle concentration. For instance it is possible to discriminate multiple tracer types that differ e.g. in the particle core size. One challenge of multi-contrast MPI is that the reconstruction problem is severely ill-posed such that in practice a perfect separation of different tracer types is not achieved. In this work, we develop a method for improving the channel separation and in turn prevent leakage from one channel into the other. Our approach exploits sparsity in both the spatial and the channel dimension. By developing a tailor regularization approach for improved multi-contrast reconstruction, we show that it is possible to significantly reduce signal leakage.