Introduction

Multi-coil receive arrays paved the way for improved diagnostic image quality by optimally combining individual coil signals to produce images with higher SNR ¹. Recent advancements have been made towards designing and efficiently manufacturing tailor-made MRI receive arrays. We have previously demonstrated body conformal, custom made receive arrays using novel fabrication methods such as screen printing, spray coating, and electroless copper ^2-4. In addition to close fit to the body, coils in receive arrays must be decoupled geometrically to prevent detuning of the individual channels. Geometric decoupling is achieved by overlapping neighboring coils to reduce mutual inductance. Having to overlap traces on body conformal coils adds another layer of complexity to its manufacturing process as additional processing steps are required to electrically connect the overlapping coils. This is especially difficult on complex curvilinear surfaces⁴.

Coaxial receive coils have been demonstrated with excellent decoupling performance without geometric decoupling^5,6. As a result, receive arrays built from Coaxial coils do not need overlapping traces to achieve adequate SNR⁵. In this work, we compare the decoupling performance of multiple transmission line topologies (Coaxial, Twisted Pair, and Microstrip) to design new 3T coil arrays. The results from this study offer novel designs for receiver coils and paves the way for future development in the area of Transmission Line Receiver Coils.

Methods

Transmission lines are specialized structures designed to transmit electromagnetic waves in a confined manner. Fields generated by transmission lines are usually enclosed, with some topologies resulting in better shielding than others. Figure 1 demonstrates how to use transmission lines as resonant coil loops. The gap on the signal conductor acts as a feed-point for the coil, while the gap on the return conductor allows for induced current to couple onto the transmission line. An equivalent circuit model is provided in Figure 2D.

Each of the three proposed transmission line topologies were fabricated with a fixed diameter of 67.5 mm. A flexible non-magnetic RG-316 cable was used for the Coaxial coils, a 28-gauge wire with Polytetrafluoroethylene (PTFE) insulation was used to form the Twisted Pairs, and Rogers RO4003C substrate was used for the Microstrip coils as shown in Figures 2A-C. The traditional coils, used as control, were fabricated on a polycarbonate substrate with copper tape as traces.

The coils were tuned and matched to 127.7 MHz with a return loss of at least 15 dB. $$$Q_{unloaded}$$$ and $$$Q_{loaded}$$$ were measured using the benchtop setup depicted in Figure 3A. Additionally, their performance was evaluated individually by measuring SNR along an image slice perpendicular to the coil. An additional decoupling experiment was performed by overlapping a pair of coils across various distances on a loading phantom and measuring insertion loss using a network analyzer (Keysight N9923A FieldFox).

Results and Discussion

To achieve body noise dominance the $$$Q_{unloaded}$$$ to $$$Q_{loaded}$$$ should be greater than two⁷. The Coaxial, Microstrip, and traditional coils all achieve body noise dominance. However, the Twisted Pair coils have a $$$Q_{unloaded}$$$ to $$$Q_{loaded}$$$ ratio of ~1.88. This low value is likely caused by the conductive losses associated with the 28 gauge wire used⁸. Thicker wire gauge could improve the $$$Q_{unloaded}$$$ performance.

Following the results from the $$$Q_{unloaded}$$$ to $$$Q_{loaded}$$$ ratio experiment, Figure 4 shows the individual performance of each coil in terms of SNR. The coils that achieve body noise dominance have similar SNR performance within the phantom. However, the Twisted Pair coil has lower SNR on the surface of the phantom. This is related to the low $$$Q_{unloaded}$$$ to $$$Q_{loaded}$$$ ratio reported by the Twisted Pair coil.

The benchtop decoupling experiment reveals adequate decoupling performance for the Coaxial and Microstrip coils. Twisted Pair and traditional coils couple heavily when the overlap is set to 0 mm. As a result, the Twisted Pair coils might benefit from geometric decoupling. At an overlap of around 20 mm, the Twisted Pair and traditional coils achieve a local minimum in decoupling performance. As reported by Roemer, geometric decoupling occurs when d (Figure 5A) is around a quarter of the coil’s diameter. For a coil diameter of 67.5 mm, geometric decoupling occurs when d is 16.875 mm. Therefore, it’s expected that coils who benefit from the geometric decoupling method described by Roemer reach a minimum in $$$S_{21}$$$ at 20 mm of overlap. From the coil topologies selected for this study, the Twisted Pair coil has the least amount of shielding. As a result, more coupling is expected between Twisted Pair elements and geometric decoupling could optimize its performance.

Conclusion

Transmission Line Receiver Coils have the potential to have better innate decoupling performance when compared to traditional coils. The results from this study show that some Transmission Line topologies offer similar single coil SNR performance to traditional coils with improved decoupling performance. Future work involves investigating other Transmission Line topologies such as Strip-Line and the Coplanar Waveguide. Additionally, the performance of these coils will be examined in multi-coil arrays.

Acknowledgements

This material is based upon work supported by the National Science Foundation Graduate Research Fellowship under Grant No. (DGE 1752814). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation. Additionally, the research was partially funded by NIH 1U01EB029427-01 and NIH 1U01EB025162-01.