Fig 1.
Design of the flexible 23-channel coil array.
a) Schematic of the rigid-flex PCB forming the array. b) Circuit diagram for a single coil element. Photographs of the custom-built array in bent (c) and flat (d) configuration. e) Photograph showing a part of the PCB assembled with lumped element components, but without preamplifiers and cables.
Table 1.
Imaging parameters for the in vivo acquisitions shown in Fig 7.
Table 2.
S-parameters of a three-element test array as function of the bending diameter.
Fig 2.
S-parameter and noise correlation matrix of the 23-channel array.
The arrangement and numbering of the individual array elements are shown on the left. The S-parameter matrix measured on the bench without preamplifier decoupling and the normalized noise correlation matrix are shown on the right. Matching better than -16 dB and decoupling better than -8 dB was achieved for all elements. Mean and maximum off-diagonal noise correlation values are 6.6% and 35%, respectively.
Fig 3.
Setup for phantom MRI experiments and SNR data.
Left: The setup for MRI experiments with the array and a spherical phantom. Right: Comparison between the developed array (FLX) and the reference coil (REF) in terms of SNR with GRAPPA acceleration factors R = 1 (no acceleration), 2, 3, and 4. Maximum g-factors in an elliptical ROI representing the occipital lope are given, and corresponding SNR-ratio maps are shown in the bottom row.
Fig 4.
Brain images acquired with the developed 23-channel array in comparison to a 20-channel reference coil.
The same image windowing and color scales were used for both compared arrays.
Fig 5.
Saggital gradient echo (GRE) images with corresponding SNR- and g-factor maps comparing the developed 23-channel array to a 4-channel flexible reference coil.
R is the applied GRAPPA acceleration factor; PE is the phase encoding direction. The same image windowing and color scales were used for both compared arrays. The sketch in the lower left corner shows the positioning of the coil on the phantom.
Fig 6.
Transversal gradient echo (GRE) images with corresponding SNR- and g-factor maps comparing the developed 23-channel array to a 4-channel flexible reference coil.
R is the applied GRAPPA acceleration factor; PE is the phase encoding direction. The same image windowing and color scales were used for both compared arrays. The sketch in the lower left corner shows the positioning of the coil on the phantom.
Fig 7.
MR images showing image quality and versatility of the flexible coil array. a) occipital cortex, b) ankle, c) hand, d) knee in axial (left), sagittal (top right), and coronal (bottom right) view, e) shoulder in axial (left) and coronal (right) view. The corresponding imaging parameters are listed in Table 1.
Fig 8.
In vivo gradient echo (GRE) images with corresponding SNR- and g-factor maps of brain (a) and knee (b).
R is the applied GRAPPA acceleration factor; PE is the phase encoding direction. The sketches on the left corner show the positioning of the coil on the subject. As indicated in the sketch, the array does not fully surround the volunteer’s knee, which is the reason for the signal drop-off in the medial region.