Highly decoupled shielded loop coils as receive array elements for 7T MRI
Irena Zivkovic1, Thomas Ruytenberg1, and Andrew Webb1

1Radiology Department, C.J. Gorter Center for High Field MRI, Leiden, Netherlands


We propose a shielded loop coil design for operation as a receive array element at 7T. The proposed coil geometry provides high decoupling between adjacent and non-adjacent loop elements without overlapping or preamplifier decoupling. A four-coil receive array was constructed for high resolution imaging of the knee at 7T. The coil coupling between the elements for in-vivo measurements was lower than -26 dB. Images with an isotropic resolution of 0.7 x 0.7 x 0.7 mm were acquired in five minutes.


The advantages of using receive arrays in MRI instead of single elements are increased SNR and being able to perform parallel imaging. Surface loop coils are common elements in such receive arrays. The major challenge is to overcome the strong coupling between the elements when multiple surface loop elements are positioned closely in an array1. Neighbouring coils can be decoupled by geometrical overlapping, via isolation networks2, or based on non-even distribution of electrical impedances around the length of the loop3, in addition to the intrinsic decoupling associated with connecting low input impedance pre-amplifiers to all coils with appropriate matching networks1.

Recently, high impedance coils have been proposed as a general method of achieving a configuration-independent high degree of decoupling: results have been shown at 3T 4. In this decoupling method there is no need to distribute lumped elements along the length of the loop. The drawback of this method is that the maximum loop size decreases as a function of static field. At 7T, the maximum diameter of the high impedance loop was noted to be 40 mm4.

In this abstract, we propose a modified high impedance coil design, adapted from the principles based on ham radio antennas5, achieving a loop diameter of 80 mm at 7T, thus extending the concept to be able to accommodate much larger loops. The proposed coil geometry provides high decoupling between adjacent and non-adjacent elements without overlapping or preamplifier decoupling. The receive performance of a four-element array specifically designed for imaging the knee was investigated in healthy volunteers.


The loop was made of coaxial cable (G 01132-06, Huber+Suhner, Switzerland) with two capacitors of 15 pF attached in series and one 27 pF capacitor connected in parallel. Two PIN diodes (MA4P7441F-1091T, MACOM, USA) were used for detuning purposes and were connected between the inner conductor of the coaxial cable and the shield, opposite from the feed point. The cable shield was interrupted at the top part of a loop (opposite from the feed point) while the inner conductor was interrupted at the bottom part5 . For comparison purposes standard capacitive-segmented loop coils were constructed with the same diameter.

All measurements were performed on a 7T Philips Achieva scanner. For both phantom and in vivo measurements a quadrature high-pass birdcage coil (Nova Medical) was used to transmit and the four channel array of the shielded loop coils to receive. A cylindrical phantom (diameter 120 mm) filled with saline solution containing 2.5 g NaCl per litre of water (ɛr=78 and σ=1 S/m) was used. In vivo measurements were performed on a knee of a healthy volunteer. In vivo images were obtained with a 3D T1-weighted GRE sequence with the following parameters: TR/TE = 5.8/2.5 ms, FA=10°, voxel size = 0.7 x 0.7 x 0.7 mm, FoV = 200 x 140 x 40 mm, two signal averages.


Figure 1 shows signal intensity maps of the four individual channels when placed on the cylindrical phantom with the coils placed directly next to one another (spacing < 1 mm). Coil coupling can be quantified by the measured noise correlation matrices, also shown in Figure 1. Decoupling better than -34 dB was measured for all channels. Decoupling better than -26 dB was obtained when the coils were overlapped by 25%, except for a channels 3 and 4 where decoupling was -16 dB. For comparison, the conventional loops of the same diameter had decoupling of -9.6 dB (when placed directly next to one another) and -14 dB with 25% overlap. Figure 2 shows the measurement setup with four receive loops placed around a knee of a volunteer. The coil coupling between each of the elements in-vivo was lower than -26 dB. Figure 3 shows high resolution gradient-echo knee images obtained in both sagittal and axial orientations.


We have shown a very simple method for constructing surface coil arrays with a high degree of inter-element decoupling. In this design there is no need for distributed lumped elements within the loop. The coils can be overlapped to various degrees with very little variation in the coupling. The concept is very similar to designs proposed in5 for low frequency amateur radio communications. The geometry also has parallels in recently-described high impedance coils4, with the difference being that we have capacitor connected in parallel to the loop while the high impedance coil uses an inductor. This difference allows construction of a coil for operation at 7T with an increased diameter.


A receive array of shielded loops shows a high degree of intrinsic inter-element decoupling, almost irrespective of the degree of overlap. This should allow the simple construction of flexible multi-element arrays for high field MRI.


This work was funded by NWO grant #13783 and ERC NOMA-MRI 670629


  1. Roemer PB, Edelstein WA, Hayes CE, Souza SP, Mueller OM. The NMR phased array. Magn Reson Med. 1990;192-225.

  2. Avdievich NI, Pan JW and Hetherington HP. Resonant inductive decoupling (RID) for transceiver arrays to compensate for both reactive and resistive components of the mutual impedance. NMR. Biomed. 2013;26(11):1547–1554.

  3. Yan X, Gore JC, Grissom AW. Self-decoupled radiofrequency coils for magnetic resonance imaging. Nat. Comm. 2018;9:3481.

  4. Zhang B, Sodickson DK, Cloos MA. A high-impedance detector-array glove for magnetic resonance imaging of the hand. Nat. Biomed. Eng. 2018;2:570-577.

  5. Demaw D. On ground low-noise receiving antennas. QST. 1988;4:30-34.


Figure 1. (top) Images from the individual channels The array elements were placed around a cylindrical phantom with no overlap between neighbouring elements. (Bottom) Measured noise correlation matrices on a cylindrical phantom with elements placed immediately adjacent to each other (left) and with 25% overlap (right).

Figure 2. (Left) Photograph of the in vivo measurement setup consisting of four non-overlapped receive loops (an outer birdcage coil was used for transmit and is not shown here). In vivo measured noise correlation matrix (right). All coupling coefficients of the receive channels were below -26 dB.

Figure 3. In-vivo images of the knee obtained using the four-element shielded loop array. A zoomed image of the cartilage is shown to demonstrate the fine structure.

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