Alexandria Engineering Journal (Oct 2022)

Selection field generation using permanent magnets and electromagnets for a magnetic particle imaging scanner

  • M. Irfan,
  • O. Mercan Dogan,
  • N. Dogan,
  • A. Bingolbali

Journal volume & issue
Vol. 61, no. 10
pp. 7685 – 7696

Abstract

Read online

This paper presents a comprehensive study on modeling and implementation of the selection field using NdFeB permanent magnets, electromagnetic coils, and a hybrid system for magnetic particle imaging (MPI). Selection fields were designed for 71 mm internal workspace and their size was based on an 11.25% enhanced ratio of Maxwell configuration. The magnetic particle imaging technique uses a gradient field to define a field of view (FOV) and generate a field-free point (FFP) which is applied to localize superparamagnetic nanoparticles (SPIONs). This study aims to develop a selection field for a new 3D magnetic particle imaging scanner at (−2.15, −2.15, 4.3) T/m gradient field in (x, y, z) directions, respectively. Three different topologies for gradient fields were numerically modeled in COMSOL and analytically calculated in MATLAB for magnetic flux density (mT) and gradient field strength (T/m). A selection field with NdFeB permanent magnets was practically implemented and its result was compared with the outcomes of a numerical simulation model. Experimental results of permanent magnets selection field topology agreed well with the numerical results from COMSOL. Numerical and analytical results of other selection field topologies were in good agreement. Spatial homogeneity of the selection field topologies was investigated over a range of 40 mm symmetric across the FFP and cost-effective analysis was performed to choose an optimum topology. A hybrid system resulted in better homogeneity over the permanent magnet and electromagnetic topologies with 96.8% spatial homogeneity, and 0.30% relative gradient field strength error. High spatial homogeneity of the gradient field minimizes the image artifacts of the MPI.

Keywords