Visualization and Localization of Implanted Devices with Parallel Transmit Array Using Reversed RF Polarization
Parnian Zarghamravanbakhsh1, John M Pauly1, and Greig Scott1

1Electrical Engineering, Stanford University, Stanford, CA, United States

Synopsis

The radiofrequency (RF) transmit field can induce current in implanted devices; therefore, it is essential to detect and minimize coupling to stimulator leads and guide-wire structures. Reverse polarization has been proposed as low-RF-power method to safely detect current in the implanted devices using birdcage coil. The purpose of this study is to demonstrate feasibility of combining knowledge of coil current and location with reverse polarization method using parallel transmit array to detect and localize implanted wires.

Introduction

A key goal in RF safety in MRI is to identify and minimize coupling to stimulator leads and guide-wire structures. The reverse polarization method has been introduced as a low power method to visualize implanted wires1. Moreover, the contrast of a reverse polarized signal can be improved by using pre-spoiler gradients2. Conceptually, wire structures act as polarization transformers through generation of linear polarized scattering fields. Reverse polarization was originally intended for birdcage coils, which produce approximately uniform RF fields. In contrast, multichannel transmitters have non-uniform transmit sensitivities, and obtaining ideal forward or reverse polarization is a challenging issue. Prior knowledge of B1+ maps is required to determine complex weights of each transmit element to synthesize forward or reverse transmission in any volume of interest. In this study, we propose transmit coil locators3,4 and RF coil current sensing5 to enable reverse polarized transmission synthesis for transceiver arrays, and demonstrate localization of coupled structures inside the excitation volume.

Methods

The study was performed on GE 1.5 T scanner with a cylindrical four-channel transmit/receive array controlled by a 4-ch Medusa console. Each coil had an RF current sensor to measure circulating currents. Each sensor modulated a high speed LED that was coupled to plastic optical fiber, and demodulated by photonic receivers. Furthermore, bidirectional couplers sensed forward and reverse power on each channel. A switch matrix selected one of current, forward V or reverse V on each channel for digitization during transmit. Three fluorine fiducial markers were placed on conductor edges of one of the coils to localize the transmit array3, and then the simulated relative complex B1+ maps for each coil was estimated using FDTD-based analysis as part of our previously proposed method4. The B1+ estimates determined the relative complex weights for each transmit channel to obtain forward or reverse polarized transmit fields in the selected volume of interest.Next, the transmit array system was calibrated to ensure the calculated voltages were applied to each transmit coil using a two-part process: 1) Making a lookup-table for correcting amplitude and phase non-linearity of each amplifier using forward power data measured by bidirectional couplers. 2) Calculating the transmit coil decoupling matrix using current sensor 6 .

All experiments were performed on a saline phantom containing an insulated wire. Projection imaging was performed using a GRE sequence with non-selective slice gradient. Separate images were acquired for forward and reverse polarized transmit fields. The reverse polarization experiment was also tested with pre-spoiler gradients to further suppress the background signal near the wire. Moreover, taking advantage of the high spatial frequency in the unspoiled reverse polarization image, a Canny edge detection filter was applied to this image in MATLAB to detect and localize the wire within the excitation volume.

Results

Figure 1 shows sagittal projection images of the wire phantom for forward, unspoiled reverse, and spoiled reverse transmit mode. As is shown in Figure 1b, the unspoiled reverse polarized transmit mode reduces background signal in selected regions near the wire. Using a gradient pre-spoiler with the reversed polarization further reduces background signal, and the wire signal is clearly distinguished in Figure 1c. Figure 2 shows sagittal and coronal projection images of the wire phantom along with the edge detection images. The location of the wire within the excitation volume is completely detected in the edge images (Figure 2c and d).

Discussion and Conclusion

In this work, the feasibility of extending forward and reverse polarization to transceiver array localization and detection of implanted wires is demonstrated. Complete coil locators and sensors are essential for rapid synthesis of the reverse polarized array modes. Pre-spoiler gradients combined with reverse polarization can significantly suppress the background signal even more. Edge detection filters can then extract a wire-line model from each projection to form a 3 dimensional model. This opens the possibility for automated computer detection, and the next stage of quantifying the level of coupling at specific wire locations, with the ultimate goal to then minimize general RF coupling by transmit array null-mode imaging7.

Acknowledgements

Grant support: R01EB019241, R01EB008108, P01CA159992, NIH T32 HL007846, and GE research support

References

1. Overall WR, Pauly JM, Stang PP, Scott GC. Ensuring safety of implanted devices under MRI using reversed RF polarization. Magn Reson Med 2010;64:823–833.

2. Overall WR, Stang PP, Pauly JM, Scott GC. Safely Detecting Device Coupling using Reversed RF Polarization and Pre-Spoiled EPI. In Proceedings of the 18th Annual Meeting of ISMRM, Stockholm, Sweden, 2010. p. 775.

3. Zarghamravanbakhsh P, Ellenor C, Pauly J, Scott G. Toward mapping using coil locators. In Proceedings of the 22nd Annual Meeting of ISMRM, Milan, Italy, 2014. p.1457.

4. Zarghamravanbakhsh P, Pauly J, Scott G. Fast 3D Algorithm for Coil Localization as an Aid in Estimation of Distribution. In Proceedings of the 23rd Annual Meeting of ISMRM, Toronto, Ontario, Canada, 2015. p. 2378.

5. Stang P, Zanchi M, Grissom W, Kerr A, Pauly J, Scott G. RF sensor considerations for input predistortion correction of transmit arrays. In Proceedings of the 18th Annual Meeting of ISMRM, Stockholm, Sweden, 2010. p. 44.

6. Scott GC, Stang P, Overall W, Kerr A, Pauly J. General Signal Vector Decoupling for Transmit Arrays. In Proceedings of the 16th Annual Meeting of ISMRM, Toronto, Ontario, Canada, 2008. p. 146.

7. Etezadi-Amoli M, Stang P, Kerr A, Pauly J, Scott G. Controlling radiofrequency-induced currents in guidewires using parallel transmit. Magn Reson Med 2014.

Figures

Figure 1. Sagittal projection images of wire phantom. a. Forward polarization. b. Reverse polarization without pre-spoiler gradient. c. Reverse polarization with pre-spoiler gradient .The pre-spoiler gradient suppresses the background signal

Figure 2. Wire localization. Sagittal (a) and coronal (b) projection images of wire phantom using reverse polarization.The wire and phantom boundary can be localized in 3 dimensional imaging space by applying edge detection filters to the sagittal (c) and coronal images(d).



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
2217