0644

Construction and Validation of Multi-Coil Technology for Accessible Head-Only Scanner

Sebastian Theilenberg¹, Yun Shang¹, Chathura Kumaragamage², Scott McIntyre², Terence W. Nixon², Robin A. de Graaf², and Christoph Juchem^1,3
¹Department of Biomedical Engineering, Columbia University, New York, NY, United States, ²Department of Radiology and Biomedical Engineering, Magnetic Resonance Research Center, Yale University School of Medicine, New Haven, CT, United States, ³Department of Radiology, Columbia University Medical Center, New York, NY, United States

Synopsis

As part of a multi-center effort to design and build an accessible 1.5 T head-only MRI scanner (NIH U01EB025153, PI M. Garwood), we designed and constructed a prototype multi-coil array fitting the limited available space and capable of generating linear and non-linear image encoding fields as well as strong concomitant localized B₀ shim fields. Here, we describe the construction process, and present system characteristics as well as experimental validation of its field generation capabilities utilizing a conventional 4 T MRI scanner.

Introduction

Recent attempts to improve accessibility to magnetic resonance imaging (MRI) relaxed various design principles, including historically high demands on the homogeneity of the magnet’s B₀ field. The presence of strong B₀ inhomogeneity, however, necessitates the usage of pulse sequences robust against off-resonance effects^1-3 and potentially complex B₀ shim fields. The dynamic multi-coil technique (DYNAMITE) is uniquely suited to provide both the (non-)linear image encoding fields necessary for such sequences, as well as highly localized B₀ shim fields^4,5 with the same set of coils.

Here, we present construction and validation of multi-coil array (MCA) designed specifically for a compact 1.5 T head-only scanner^6,7 whose magnet includes a patient window and excludes the subject’s shoulders to increase accessibility and enable novel motor-coordination studies. We describe the manufacturing of our prototype, its specifications, and experimentally validate its field generation capabilities as essential steps towards subsequent installation and use of this novel MCA in the 1.5 T head-only scanner.

Manufacturing

The MCA was designed as a compact system with an inner diameter of 300 mm, a radial thickness of 22 mm and an overall length of 530 mm. Four cylindrical rings of eight coils each are arranged on a cylindrical surface with the upper most ring tapered inwards to match the magnet’s geometry⁸. Two coils in the second ring are fused together to incorporate the patient window, resulting in 31 coils total. Each coil is a sandwich of an outer and an inner element of 100 turns of 20 AWG copper wire each, with PTFAN tubes for water-cooling running in the gap between both elements. Channels are driven individually with current up to 5 A.

The coil elements were individually wet wound^7,9 with epoxy and preassembled with cooling tubes into the four rings. Twisted pair lead wires were connected to all coils, 12 resistance temperature detectors (RTDs) were distributed throughout the coils, and everything was assembled inside a custom-made aluminum mold (Figure 2A). After adding woven glass fibers the setup was vacuum infused with epoxy resin.

Fiberglass flanges at the bottom and a tube at the top were added for mounting (Figure 2B), and all wires and cooling tubes were terminated at receptacles and custom-made manifolds, respectively, in a panel in the lower back of the system (Figure 2C).

Characterization

Resistances and inductances were measured individually for each channel (Table 1). The standard coils have a resistance of 1.95(1) Ω and an inductance of 3.75(2) mH, while the window coil elements and the conical coils have 1.97(6) Ω/2.62(8) mH and 1.69(1) Ω/3.08(2) mH, respectively. Connecting the system to a multi-channel amplifier system (Resonance Research Inc., Billerica, MA, USA) via a filter box added an additional 0.5 Ω and 0.2 mH to each channel. With the given specifications, and driving both window coil elements individually, the amplifier can ramp the MCA from 0 to 100% in 300 μs.

Using a custom-made gradient controller¹⁰, the system was driven with constant currents applied to all coils while connected to a water chiller providing a total flow of 6 l/min at 20 °C through the cooling tubes. Steady-state coil temperatures increased with the applied current, with the window coil and the conical coils in the top ring showing the highest temperatures (Figure 3A). Reducing the water flow to 4 l/min did not significantly change the observed temperatures (not shown). The temperature curves closely fitted a mono-exponential function as expected with time constants in the order of 4 min (Figure 3B), which is long compared to the switching times in typical MRI sequences. For that reason, the presented steady-state experiment determines the system’s heating behavior at a given duty cycle. The maximum safe coil temperature was defined as 50 °C, corresponding to a maximum constant current of 3.5 A or a duty cycle of 70% at 5 A.

Field Generation

The MCA’s field generation capability was verified in a 4 T Bruker human scanner at the Yale MR Research Center. To this end, calibration maps (in Hz/A) of the MCA’s individual coils were measured by acquiring several B₀ field maps at different current settings¹¹. This calibration was then utilized to calculate optimal coil currents for various target fields using B0DETOX^12,13.

The MCA was tested on two different targets, a honeydew melon and a pomegranate on top of seven bananas. Three MRI sequences were used, multi-slice spin echo, 3D gradient echo and a recently presented low-angle-combined-echo sequence³ (Figure 4, image parameters in caption). All images were acquired using the scanner’s regular gradients system as well as using MC generated gradients only. Even though the imperfections of the generated fields are known, no image distortion correction was applied here.

Discussion & Outlook

We successfully designed and manufactured a prototype MCA for a novel accessible head-only MR scanner. The final system specifications closely match the design targets and predictions, with each channel being able to be driven with 5 A maximum by the designated multi-channel amplifier. The system’s high maximum duty cycle enables the generation of both image encoding and concomitant shim fields with the full dynamic range of the system.

System integration with the overall accessible MR system is planned for early 2022.

Acknowledgements

This research was supported by the National Institutes of Biomedical Engineering & Bioengineering of the National Institutes of Health under award number U01EB025153.

References

1. Patz S, Wong STS, Roos MS. Missing pulse steady‐state free precession. Magn Reson Med. 1989;10(2):194–209.
2. Kobayashi N, Parkinson B, Idiyatullin D, Adriany G, Theilenberg S, Juchem C, et al. Development and validation of 3D MP‐SSFP to enable MRI in inhomogeneous magnetic fields. Magn Reson Med. 2021;85(2):831–44.
3. Theilenberg S, Kumaragamage C, McIntyre S, Nixon TW, Juchem C, De Graaf RA. Low-Angle Combined-Echo (LACE) Imaging in Highly Inhomogeneous B0 Magnetic Fields. Proc Intl Soc Magn Reson Med. 2021;29:0004.
4. Juchem C, Nixon TW, Diduch P, Rothman DL, Starewicz P, de Graaf RA. Dynamic shimming of the human brain at 7 T. Concepts Magn Reson Part B Magn Reson Eng. 2010;37B(3):116–28.
5. Juchem C, Mullen M, Kumaragamage C, DelaBarre L, Adriany G, Brown PB, et al. Dynamic Multi-Coil Technique (DYNAMITE) MRI on Human Brain. Proc Intl Soc Magn Reson Med. 2019;27:0219.
6. Garwood M, Mullen M, Kobayashi N, DelaBarre L, Suddarth S, Idiyatullin D, et al. A compact vertical 1.5T human head scanner with shoulders outside the bore and window for studying motor coordination. Proc Intl Soc Magn Reson Med. 2020;28:1278.
7. Juchem C, Theilenberg S, Shang Y, Ghazouani J. Magnetic Resonance Imager with Coils for Arbitrary Image Space and Fabrication Thereof. US Patent WO2021/022252, February 4, 2021.
8. Theilenberg S, Shang Y, Kobayashi N, Parkinson BJ, Juchem C. Multi-Coil Array for Combined Imaging and B0 Shimming in a Portable Head-Only Scanner. Proc Intl Soc Magn Reson Med. 2019;27:1480.
9. Shang Y, Theilenberg S, Ghazouani J, Juchem C. A Coil-Winding Machine for Wet-Winding of Resin-Impregnated Curved Wire Patterns. Proc Intl Soc Magn Reson Med. 2020;28:4248.
10. Nixon TW, McIntyre S, de Graaf RA. The Design and Implementation of a 64 Channel Arbitrary Gradient Waveform Controller. 2017;25:0969.
11. de Graaf RA, Juchem C. CHAPTER 4. B0 Shimming Technology. In: Webb A, editor. Magnetic Resonance Technology: Hardware and System Component Design. Cambridge, UK: The Royal Society of Chemistry; 2016:166–207.
12. Juchem C. B0DETOX - B0 Detoxification Software for Magnetic Field Shimming. Columbia University Technology Ventures; 2017. Available from: http://innovation.columbia.edu/technologies/cu17326_b0detox
13. Juchem C, Herman P, Sanganahalli BG, Brown PB, Mcintyre S, Nixon TW, et al. DYNAmic Multi-coIl TEchnique (DYNAMITE) shimming of the rat brain at 11.7T. NMR Biomed. 2014;27(8):897–906.

Figures

Figure 1: CAD model of the accessible head-only scanner and the Multi-Coil Array. The MCA is located inside the magnet bore (white), sandwiched between the passive shim cassette on the outside and the RF coil on the inside (not shown). All components including the magnet itself are located above the subject’s shoulders and include a patient window, enabling free limb movement.

Figure 2: MCA hardware construction. A) View of the back of the coils assembled on the inner part of infusion mold, including lead wires and cooling tubes prior to the epoxy infusion. The cable ties were temporarily used to hold all components in place. Subsequently, the assembly was covered with woven glass fiber as seen in the back of the coils, and the mold was closed. B) Resultant Multi-Coil Array after vacuum epoxy infusion and with glued on mounting flanges. C) Bottom view showing wire looms, receptacles, and manifolds mounted to a 3D-printed panel fixated below the back flange.

Figure 3: Coil temperatures while driving the MCA with constant currents. A) Steady-state coil temperatures as measured by the RTDs at different DC currents applied to all coils. Individual RTDs are color coded by their location from bottom (blue) to the top ring (yellow). RTD 4 is at the window coil. A maximum temperature of 50 °C was deemed save and can be maintained at continous 3.5 A. B) Exemplary temperature progression as measured by one RTD during heating with 4.25 A DC and subsequent cooling without current.

Table 1: System characterization. Resistance and Inductance were measured for each channel individually using an LCR meter. Coils 9a/b indicate the inner and outer elements of the larger window coil, channels 24-31 indicate the tilted coils in the top row of the system. Cables to the amplifier and filters added additional 0.5 Ω and 0.2 mH to each channel.

Figure 4: MR Imaging using MC generated fields. Shown are results for three imaging sequences (a 3D gradient echo sequence (GRE, matrix 256x256x256, FoV 256x256x75 mm³), a multi-slice spin-echo sequence (SE, matrix 256x256x25, FoV 256x256x75 mm³), and a low-angle-combined-echo sequence (LACE, matrix 128x128x32, FoV 256x256x32 mm³). Shown are central axial slices through a honeydew melon (left two columns) and a pomegranate on top of a couple of bananas (right two columns), each imaged using the scanners gradient system (Gradient) and MC generated fields (MC).

Proc. Intl. Soc. Mag. Reson. Med. 30 (2022)

0644

DOI: https://doi.org/10.58530/2022/0644