A Novel Metamaterial-Inspired RF-coil for Preclinical Dual-Nuclei MRI

Abstract

In this paper, we propose, design and test a new dual-nuclei RF-coil inspired by wire metamaterial structures. The coil operates as a result of resonant excitation of hybridized eigenmodes in multimode flat periodic structures comprising several coupled thin metal strips. It was shown that the field distribution of the coil (i.e. penetration depth) can be controlled independently at two different Larmor frequencies by selecting a proper eigenmode in each of two mutually orthogonal periodic structures. The proposed coil requires no lumped capacitors to be tuned and matched. In order to demonstrate the performance of the new design, an experimental preclinical coil for ¹⁹F/¹H imaging of small animals at 7.05T was engineered and tested on a homogeneous liquid phantom and in-vivo. The results demonstrate that the coil was both well tuned and matched at two Larmor frequencies and allowed image acquisition at both nuclei. In an in-vivo experiment, it was shown that without retuning the setup it was subsequently possible to obtain anatomical ¹H images of a mouse under anesthesia with ¹⁹F images of a tiny tube filled with a fluorine-containing liquid and attached to the body of the mouse.

Figure 1

Proposed dual-nuclei RF-coil, comprising loop feed and two wire metamaterial-inspired resonators, with subject inside RF-shield of MRI (a); illustration of miniaturization principle for resonator of two metal wires: current distributions and equivalent circuits of resonators with two half-wavelength (b) and shortened (c) parallel wires.

Figure 2

Geometry and eigenmode H-field patterns (normal component with respect to plane of strips), in a.u.: long-wire resonator (a,c,e,g,i,k) and short-wire resonator (b,d,f,h,j,l).

Figure 3

Eigenmode transverse plane H-field patterns, in a.u.: long-wire resonator (a–e) and short-wire resonator (f–j).

Figure 4

Normalized magnetic field magnitude of eigenmodes of the orders 1–5 depending on distance away from plane of strips: (a) long-wire resonator; (b) short-wire resonator.

Figure 5

Simulated and measured values of reflection coefficient S₁₁ of proposed coil for ¹H/¹⁹F imaging vs. frequency (a); simulated S₁₁ vs. frequency of proposed coil with variable phantom size for ¹H channel (b) and ¹⁹F channel (c).

Figure 6

Simulated normal magnetic field component (a.u) in vicinity of RF-coil, operating using modes 1,3 (a,c) and modes 1,1 (e,j) of short and long wire resonator respectively: 282.6 MHz (a,e) and 300.1 MHz (c,j); simulated distributions of $|B_1^{+}|$ for 0.5 W accepted power, $\mu \mathrm{T}/\sqrt{\mathrm{W}}$, in central transverse plane of RF-coil operating using modes 1,3 (b,d) and modes 1,1 (f,h) of short and long wire resonator respectively: 282.6 MHz (b,f) and 300.1 MHz (d,h).

Figure 7

Reconstructed images acquired with Bruker PharmaScan 7 T of cylindrical liquid phantom (mixture of 60% 2-2-trifluorethanol and 40% water) with gradient echo sequence: ¹⁹F (a); ¹H (b); normalized image profiles vs. depth to phantom in central transverse plane (c); normalized image profiles vs. axial distance along central axis of phantom (20 mm depth) (d).

Figure 8

In-vivo setup including mouse under anesthesia and syringe containing fluorine compound (a) and anatomic gradient echo coronal-plane image of mouse acquired using the manufactured dual-nuclei coil (b).

Figure 9

In-vivo anatomic gradient echo axial-plane images of mouse and ¹⁹F images of syringe with a mixture of 60% 2-2-trifluoroethanol and 40% water attached to mouse body in fourteen different slices acquired using proposed dual-nuclei coil.

Figure 10

Experimental RF-coil: (a) PCB parts including loop feed, short-wire resonator and long-wire resonator and their assembly with spacer, phantom and RF-cable; (b) bed parts: holder for resonators and for small animal.