0438

An eight-channel array coil for zero echo time imaging

Manuela Barbara Rösler¹, Markus Weiger¹, David Otto Brunner¹, Romain Froidevaux¹, Thomas Schmid¹, and Klaas Paul Pruessmann¹

¹ETH Zurich and University of Zurich, Zürich, Switzerland

Synopsis

With dedicated short-T2 techniques such as zero echo time (ZTE) imaging not only rapidly decaying signal from the sample but also from the RF coil (plastic housing, glue, solder flux etc.) is detected. To enable experimental examination of parallel imaging opportunities, a 1H-reduced eight channel array was designed. ZTE images of short-T2* rubber samples and PETRA images of human head could be acquired.

INTRODUCTION:

ZTE acquisitions are used to capture fast relaxing MR signal not visible in images acquired with ordinary sequences. Because of the high short-T2 sensitivity of these technique, images can be contaminated by signal form plastic housings, cable isolations, glues, epoxies, solder flux etc., in particular form the RF coil (1). To avoid related image artifacts, loop and birdcage coils containing almost no 1H were developed (2-4).

In this abstract, an eight-channel transmit-receive coil with minimized 1H-background signal is presented to examine experimentally parallel imaging opportunities (5) for ZTE sequences. The array is optimized to a high-performance gradient insert (6) used for high-resolution imaging of fast relaxing spins (7).

METHODS:

The coil was built up using cupper tubes with an outer diameter of 2 mm and is supported by PTFE holders with fitting holes (Figure 1) fixed to each other with brass screws. With two low 1H-signal trim capacitors (NMNT-Series, Voltronics) framed by PTFE supports, the eight loop elements can be tuned and matched to different loads e.g. human heads, knees and feet and for testing purposes to the unloaded case. Electrical isolation between the subject and the coil electronics is ensured by a glass cylinder with an inner diameter of 23 cm. The shielding meshed conductive textile (2) is hold by another glass cylinder. Cable traps are included into the coil housing and fabricated from semi-rigid cables containing PTFE as a dielectric (EZ_141_CU_TP, Huber+Suhner). The windings of the cable traps and of the decoupling transformers made from silver plated cupper wires with a diameter of 1.2 are isolated by surrounding PTEF tubes (Figure 1 b+c). After mounting, the coil was cleaned with acetone. The exact electrical design is described in (8) .

Initial measurements were performed on a clinical 3T whole body scanner equipped with a high performance gradient insert (6). Symmetrically biased T/R switches (9) optimized to 128 MHz enable ZTE data acquisition with a customer-built spectrometer (10) as described in (11). RF power provided by the amplifier was split passively by a butler matrix.

Figure 2a shows the array coil placed inside the gradient. From stacked rubbers (Figure 2b) with T2* ≈ 300 μs, multi-channel ZTE images were acquired and reconstructed in two ways: 1) Algebraic ZTE reconstruction was performed on each channel separately (12) and then combined using sum-of-squares calculation without utilizing the encoding capabilities of the coil sensitivities. 2) Data from all channels was used together in one ZTE-SENSE reconstruction, thus encoding capabilities of the array is employed to improve conditioning of the reconstruction problem suffering from the dead time gap in central k-space. Sensitivity maps were obtained from a ZTE data set acquired with small gap size. In vivo performance were demonstrated with PETRA imaging of a human head. Block pulse length was 2 µs and repetition time 1 ms in all imaging experiments.

RESULTS:

Figure 3: The array itself has low MR active content. Figure 4: ZTE-SENSE reconstruction improves image quality at moderate gap sizes in comparison to ordinary algebraic reconstruction. Figure 5: In vivo head imaging capturing short-T2 signal is feasible with the designed array.

DISCUSSION:

Initial ZTE images of a short-T2 sample and human head acquired with an array coil are presented. SNR gain from array reception might be reduced by additional losses due to coupling in the transmit mode since usually the Ernst angle is not reached with short RF pulses required in ZTE imaging and SNR efficiency is compromised by too small flip angles.

First example of ZTE-SENSE reconstruction is promising, but requires more detailed investigations. Exact influence of the improved conditioning of the algebraic reconstructions on acceptable gap sizes and possible angular undersampling still remains to be evaluated. With the developed hard- and software this can be studied.

Acknowledgements

Thanks to Stephen Wheeler and Stephan Siegrist for building up the mechanical housing of the coil.

References

1. Horch RA, Wilkens K, Gochberg DF, Does MD. RF coil considerations for short-T2 MRI. Magnetic Resonance in Medicine 2010;64(6):1652-1657.

2. Weiger M, Brunner DO, Schmid T, Froidevaux R, Rösler MB, Gross S, Pruessmann KP. A virtually 1H‐free birdcage coil for zero echo time MRI without background signal. Magnetic resonance in medicine 2017;78(1):399-407.

3. Rösler MB, Weiger M, Brunner DO, Schmid T, Froidevaux R, Pruessmann KP. An RF Birdcage Coil Designed for an Insert Gradient Coil Dedicated to Short-T2 MRI. ISMRM 25th Annual Meeting. Honolulu2017. p 2668

4. Rösler MB, Weiger M, Schmid T, Brunner DO, Froidevaux R, Pruessmann KP. Ultrasonic soldering on glass for the construction of MRI coils with minimized background signal in short-T2 images. ESMRMB 2016; Vienna.

5. Oberhammer T, Weiger M, Hennel F, Pruessmann KP. Prospects of Parallel ZTE Imaging 2011; Montreal. p 2890.

6. Weiger M, Overweg J, Rösler MB, Froidevaux R, Hennel F, Wilm BJ, Penn A, Sturzenegger U, Schuth W, Mathlener M, Borgo M, Börnert P, Leussler C, Luechinger R, Dietrich BE, Reber J, Brunner DO, Schmid T, Vionnet L, Pruessmann KP. A high-performance gradient insert for rapid and short-T2 imaging at full duty cycle. Magnetic Resonance in Medicine 2018;79(6):3256-3266.

7. Froidevaux R, Weiger M, Rosler M, Wilm B, Hennel F, Luechinger R, Dietrich B, Reber J, Pruessmann KP. Ultra-High-Bandwidth, High-Resolution MRI of Fast Relaxing Spins. ISMRM 25th Annual Meeting. Honolulu2017. p 4037

8. Rösler MB, Leussler C, Brunner DO, Schmid T, Luechinger RC, Weiger M, Hennel F, Pruessmann KP. A head transmit-receive array for a high performance gradient insert. submitted to ISMRM 2019

9. Brunner DO, Furrer L, Weiger M, Baumberger W, Schmid T, Reber J, Dietrich BE, Wilm BJ, Froidevaux R, Pruessmann KP. Symmetrically biased T/R switches for NMR and MRI with microsecond dead time. Journal of Magnetic Resonance 2016;263:147-155.

10. Dietrich BE, Brunner DO, Wilm BJ, Barmet C, Gross S, Kasper L, Haeberlin M, Schmid T, Vannesjo SJ, Pruessmann KP. A field camera for MR sequence monitoring and system analysis. Magnetic Resonance in Medicine 2016;75(4):1831-1840.

11. Weiger M, Brunner DO, Dietrich BE, Müller CF, Pruessmann KP. ZTE imaging in humans. Magnetic Resonance in Medicine 2013;70(2):328-332. 12. Weiger M, Pruessmann KP. MRI with Zero Echo Time. eMagRes: John Wiley & Sons, Ltd; 2007.

Figures

a) Picture of the ZTE array coil in a plastic holder not used inside the scanner. b) Transformers adjusted to decouple neighboring elements. c) Cable trap and trimmer capacitors.

a) Array coil placed inside the gradient insert with phantom built up from rubbers (T2* ≈ 300 µs) shown in b) placed on a signal-free PTFE holder.

PETRA image of the empty coil (FOV = (400 mm)^3; isotropic resolution = 3.12 mm; bandwidth = 500 kHz; gap = 7.8 Nyquist dwells filled with single shot acquisitions) averaged over 15 slices to visualize residual 1H background signal. Close to the conductors of the coil slight signals are visible. Highest signal intensity was caused most likely by soldering flux flown inside one of the PTFE tubed surrounding the transformer. Because of that, all following image were acquired with a field of view encoding the coil structures.

a) Sum-of-squares ZTE image (FOV=(330mm)^3;isotropic resolution=2mm;bandwidth=150kHz) with a gap size of 1.5 Nyquist dwells calculated from the data of the eight channels reconstructed independently showing the stack of rubbers (Figure 2). b) Gap size was increased artificially to 3.5 Nyquist dwells and the same reconstruction method as in a) was applied. Artifacts due to the ill-posed problem are visible. c) ZTE-Sense reconstruction of the same dataset as in b) yield to a significant reduction of artifacts. d) The masked used in the sense reconstruction was applied to b) to exclude that better image quality was only caused by masking.

In vivo PETRA images of human head with acquired with an array coil (FOV = 330 mm)^3; isotropic resolution = 1.47mm; bandwidth = 250 kHz; gap size = 4 Nyquist dwells). No gradient geometry correction and no intensity correction was performed.

Proc. Intl. Soc. Mag. Reson. Med. 27 (2019)

0438