The monolithic design of self-resonant transmission line resonators (TLRs) ¹ is attractive for developing small-sized coils that can be made flexible or fabricated with superconducting material and thus offers several strategies for improving the SNR in NMR experiments ². In a previous work ³, we investigated the performance of contactless tuning and matching techniques based on electric and inductive coupling and we evaluated on the bench the feasibility of implementing these techniques using an automation system based on piezoelectric actuator. In the present work, we have implemented a piezo-motor based automation system for imaging experiments at 4.7 T and we present the first image acquired with a TLR being automatically matched.

This work was conducted with a 6 turn – 4 gap TLR 4 tuned at 199.8 MHz (¹H at 4.7 T) by dielectric coupling with a FR-4(Glass-enforced epoxy, ε_r = 4.6) slice. The TLR was inductively coupled to a pickup-loop attached to a piezoelectric motor (PiezoLEGS®, PiezoMotor, Sweden). The matching procedure was performed using the automation system so as to adjust the horizontal distance between the pickup-loop and the coil. The automatic system consists of a graphical user interface (GUI) developed with MATLAB (MathWorks, Massachusetts, USA) and a driver board for controlling the piezoelectric motor and communicating with the computer. The TLR was successfully matched to more than -30 dB inside the bore of the magnet using the set-up shown in the figure 1.

MRI was done on a home-assembled 4.7 T scanner controlled by an Apollo sequencer (Tecmag, Texas USA) equipped with a BGA12 gradient bore (400 mT/m, Bruker Biospin, Germany). The parameters of the 3DGE acquisitions were TR/TE 29.78/2.57 ms, FOV (6 cm * 6 cm * 6 cm) and pixel bandwidth 398 Hz. The coil was loaded with a rectangular box-shaped phantom (5.9 cm * 4.2 cm * 6 cm) filled with water. The TLR was fixed to the bottom of a 6 mm thick Plexiglass support with the sample placed on top of it (Figure .1). In order to evaluate the influence of the piezoelectric actuator on the image quality, we acquired a first image using the piezo-motor of the automation system only, placed 7 cm horizontally away from the center of the TLR (Figure 1). We acquired a second image in the same condition than the first one but adding another piezo-motor fixed on a side of the sample. The later configuration corresponds to a worst case regarding the position of the piezo-motor. Both experiments were done after achieving the desired matching level using the automation system.

Figure 2 shows the image obtained in the axial plane in the presence of the piezo-motor of the automation system only. No significant artefacts were observed. Figure 3 shows the image acquired in the presence of the second piezo-motor fixed to the sample wall. It demonstrates that the close proximity of the piezoelectric motor to the sample brings inhomogeneity to the B1-field of the coil. The piezo-motor based automation system successfully achieved inductive matching of a TLR inside a 4.7 T NMR scanner and no artifacts were observed on the image. However piezoelectric motor should be placed at a minimum distance from the coil and the sample so as not to perturb the B1-field homogeneity. Since matching the TLR can be done remotely, it facilitates the matching procedure in small-sized bores. Furthermore, such a contactless automation system is of interest when the accessibility to the RF coil is limited such as using HTS TLRs in a cryogenic environment. In future work, extending the system for both automatic tuning and matching simultaneously and compensating for their mutual influence is of concern.