Anatomically adaptive local coils for MR Imaging - Evaluation of stretchable antennas at 1.5T
Bernhard Gruber1 and Stephan Zink2

1Medical Engineering - School of Applied Health and Social Sciences, University of Applied Sciences Upper Austria, Linz, Austria, 2R&D HW LC, Siemens Healthcare GmbH, Erlangen, Germany


This abstract is a first investigation on antenna materials and designs for anatomically adaptive local coils for MR Imaging. To overcome the SNR losses by poorly loaded and non-fitting RF coils, we proposed a stretchable antenna design. Each loop has the ability to reversible stretch up to 100% of its original size, to be anatomically adaptive to different shapes and sizes in three dimensions. Through bench measurements and MR Imaging at 1.5T we investigated different stretchable antenna materials, that fit the defined requirements. The results of stretchable loops showed an in average SNR loss of under 10% in comparison to standard loops, but we suppose that the improved filling factor will lead to much higher SNR of the adaptive loops. Further research may consider different improvements.


To combine the ability of high-quality receiver arrays for MRI with improved signal-to-noise ratio (SNR), usability in handling and patient comfort along with higher patient throughput, it is feasible to develop an adaptive local coil. The coil should be anatomically adaptive to different shapes and sizes in three dimensions [1]. This adaptivity causes frequency shifts of the larmor frequency, which entails in variations of the impedance that each coil presents to its connected preamplifier. Today’s coils are all non-stretchable, but many of them are flexible. To explore the properties of adjustable coils, we present pre-research development results about a stretchable 2-channel array.

Materials and Methods

The main problem that emerges from the change in coil size and shape is the frequency shift. We identified two acceptable materials: 5mm-wide AMOTAPE® conduct elast with PTFE insulated copper strands. These strands float in wavy (meander) line along the tape (100% elasticity). The 2nd material is a 5mm-wide highly ductile cooper tape (HDC) with 50µm dielectric thickness and 35µm copper thickness (DuPontTM Pyralux®), which is arranged waved within a stretchable base material. To test the material properties, we used a partially stretchable design of the coil loop (Fig. 1). The dimensions of a single loop are 100x100mm. The rigid parts where made of 5mm-wide copper tape. To test the properties of the double-loop (DL) and maintain homogeneous stretching, a test skeleton for reproducible testing was 3D printed, and four rigid loops where constructed, matched and tuned (0, 10, 20, 30% stretching; 60–90mm stretchable area). These four rigid DLs serve as standard to compare the stretch DL to. Adjacent coil elements were inductively decoupled from one another using critical overlap: The distance between the center of each element is acenter=0.9*d=90mm. Residual coupling was suppressed by preamplifier decoupling [2]. Capacitors in rigid sections were used to tune the loops to the resonance frequency. The DL is attached to preamplifiers (Siemens Healthcare, Erlangen, Germany) [3]. The measurement protocol consists of 16 measurements with four different states of each Stretch-Loop. Measurements were done at 1,5T Siemens MAGNETOM Aera (Siemens Healthcare, Erlangen, Germany). Imaging was performed in a 5mm safety-distance to the phantom-structure (Loader Phantom (per 1000g H2O dist.: 1.25g NiSO4 x 6H2O, 5g NaCl) and Bayol-Oil Phantom (per 1000 ml Bayol-Oil: 0.011g MACROLEX blue)) (Fig. 1). We used a Spin echo sequence in transverse orientation (TE=15ms, TR=300ms, FoV=300mm) with a slice thickness of 5mm and an acquisition matrix of 256x256 and a bandwidth of 130 Hz/pixel. The voxel size was 1.2x1.2mm.


Because of the stretch elements in the loop, the coil dimension can change and this causes a shift in resonance frequency. The coil coupling in the tested design is negligible for all stretching states (k*Qloaded is lower than 1). Thus, there is no need for strategies to mitigate inductive coupling [4]. The frequency shift for AMOTAPE® ranged from 61.83 to 65.67 MHz and for HDC from 61.75 to 65.69 MHz. The preamp decoupling ranged from -26.5 to -9.82dB, and the S12 decoupling between the nearest loops ranged from -21.94 to -16.3dB. (The HDC had slightly worse results.) The ESR for the HDC DL range from 4.3Ω (0% stretch) to 5.4Ω (30% stretch) and the AMOTAPE® DL from 1.66Ω (0% stretch) to 1.89Ω (30% stretch), included the mainboard with active/passive decoupling and matching circuitry and the RF-fuse. The SNR loss of the HDC ranged from 14.21 to 8.19% in 20mm depth and 11.31 to 6.55% in 50mm depth. The SNR was evaluated at a single voxel in 20 and 50mm depth from the surface in the middle transverse slice of the phantom. The AMOTAPE® loses 14.58 to 2.37% in 20mm depth and 12.50 to 5.55% in 50mm depth.


The goal of this project was to investigate the potential image quality of stretchable loops. The optimal solution in terms of adaptive adjustment would be a coil array, whose elements adjust their size, shape and configuration by stretching and thus automatically surround each target anatomy as close as possible. Different conductor materials were investigated as well as the influence of stretching on the SNR performance of single coil elements and coil arrays. In direct comparison of the stretch DLs to the standard DLs, the SNR loss is in average under 10%. We suppose that the stretch loop design used in an array leads to improved SNR because of the adaptive properties and the thereby improved filling factor. A stretchable ‘Demonstrator’- array for imaging of differently sized body parts at 1.5 T would be a next step.


No acknowledgement found.


[1] Nordmeyer-Massner et al. MRM (2012), 67, 872-879. [2] Roemer et al. MRM (1990), 16, p.192. [3] Keil B, et al. MRM (2012), 70:1, p.248-258. [4] Vester et al. Proc. ISMRM (2012), 2690.


3D-printed testing skeleton with 2 kinds of stretchable materials.

SNR comparison of stretchable Double-Loops to standard Double-Loops in (1) 20 and (2) 50mm depth.

Frequency Shift (upper diagram) and coil coupling (lower diagram) of stretch loops and rigid loops.

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