Home
*Note in this screen recording it says ~14mT, it is actually 10mT. The file will be updated shortly*
This wiki describes many components of the spectrometer we developed and the particular design process that led to these designs. The hope is that this Wiki can be a resource for those seeking to build a similar system and by discussing the design process, enable other users to make modifications to better suit different applications. Further, as one of the goals for the OS-MPI initiative is to be maximally accessible to newcomers in the field of MPI, this Wiki is written to be informative to that audience as well as more experienced members of the community. For any reader, we encourage feedback if you find points of confusion, mistakes, or missing information, as we are continually learning as the project grows.
There are many MPS systems discussed in publications (see here for a selection), and others available for purchase, each with their own strengths and drawback, yet compiling information from literature into a functioning design can be difficult (especially if the lab's focus isn't hardware). The goal of this project is to create a design for a simple, cheap, and conveniently packaged MPS system that can be constructed and implemented in any MPI or SPION lab.
The MPS system has three major divisions - a transmit (Tx) system, a receive (Rx) system, and the DC-bias system. Unlike the MPI imager we describe, this MPS device has no major moving mechanical parts or FFL magnets (the sample positioner is motorized, but the mass is very small and speed is fairly slow). With the transmit system there is a filter that has been designed to be flexible with respect to it's tuning to allow for easy switching between drive frequency ranges similar to (1). On the Rx side, we have the gradiometer coil, a home-built pre-amplifier.
As of now, there are three primary operation modes, a spectroscopy mode giving the system its name, a magnetometry mode, and a relaxometry mode. In the spectroscopy mode, the user drives particles within the Rx coil into saturation with a high amplitude AC signal and looks at the resulting frequency spectra see S. Biederer et al.,2009(2). With magnetometry, the goal is to superimpose a high amplitude near-DC signal with a lower amplitude AC field at the drive frequency; in this mode, the signal is recorded at the drive frequency with the signal amplitude being proportional to the magnetic susceptibility of the particles at that particular magnetic dc-bias. Finally, the relaxometry mode has been heavily employed in X-space reconstruction and essentially mimics the net fields on a discrete particle within an MPI's FOV by applying a high-frequency drive field at "normal imaging" amplitudes (5-20mT) and superimposing a near DC, high amplitude (>=50mT) bias field to mimic the gradient field's saturation effect. In "relaxometry mode" the non-linearities are used to elucidate the relaxation time constant of the particles. Relaxometry is described in detail by N. Garraud and colleagues in this paper. For further discussion on these modes see the modes of operation page.
The sensitivity of the system illustrated by plotting the signal amplitude from various masses of Synomag-D 70nm against the mass of each sample. The volume of each sample was 4 microliters and acquisition time was under one second for each.
Histogram comparing 500pg and 1ng to the signal with the sample out.
This shows that DI water is not causing any substantial signal and influencing the results.
(1) A. Behrends et al., Introducing a frequency-tunable magnetic particle spectrometer, 2015
(2) S. Biederer et al., Magnetization response spectroscopy of superparamagnetic nanoparticles for magnetic particle imaging, 2009
(3) N. Garraud et al., Benchtop magnetic particle relaxometer for detection, characterization and analysis of magnetic nanoparticles