Analog receive signal processing for magnetic particle imaging

Purpose: Magnetic particle imaging (MPI) applies oscillating magnetic fields to determine the distribution of magnetic nanoparticles in vivo. Using a receive coil, the change of the particle magnetization can be detected. However, the signal induced by the nanoparticles is superimposed by the direct feedthrough interference of the sinusoidal excitation field, which couples into the receive coils. As the latter is several magnitudes higher, the extraction of the particle signal from the excitation signal is a challenging task. Methods: One way to remove the interfering signal is to suppress the excitation signal by means of a band-stop filter. However, this technique removes parts of the desired particle signal, which are essential for direct reconstruction of the particle concentration. A way to recover the entire particle signal is to cancel out the excitation signal by coupling a matching cancellation signal into the receive chain. However, the suppression rates that can be achieved by signal cancellation are not as high as with the filtering method, which limits the sensitivity of this method. In order to unite the advantages of both methods, in this work the authors propose to combine the filtering and the cancellation technique. All methods were compared by measuring 10 mu l Resovist, in the same field generator only switching the signal processing parts. Results: The reconstructed time signals of the three methods, show the advantage of the proposed combination of filtering and cancellation. The method preserves the fundamental frequency and is able to detect the tracer signal at its full bandwidth even for low concentrations. Conclusions: By recovering the full particle signal the SNR can be improved and errors in the x-space reconstruction are prevented. The authors show that the combined method provides this full particle signal and makes it possible to improve image quality. (C) 2013 American Association of Physicists in Medicine. [http://dx.doi.org/10.1118/1.4794482]