Design and application of combined 8-channel transmit and 10-channel receive arrays and radiofrequency shimming for 7-T shoulder magnetic resonance imaging

Abstract

Objective: The objective of the study was to investigate the feasibility of 7-T shoulder magnetic resonance imaging by developing transmit and receive radiofrequency (RF) coil arrays and exploring RF shim methods.

Materials and methods: A mechanically flexible 8-channel transmit array and an anatomically conformable 10-channel receive array were designed and implemented. The transmit performance of various RF shim methods was assessed through local flip angle measurements in the right and left shoulders of 6 subjects. The receive performance was assessed through signal-to-noise ratio measurements using the developed 7-T coil and a baseline commercial 3-T coil.

Results: The 7-T transmit array driven with phase-coherent RF shim weights provided adequate B₁⁺ efficiency and uniformity for turbo spin echo shoulder imaging. B₁⁺ twisting that is characteristic of high-field loop coils necessitates distinct RF shim weights in the right and left shoulders. The 7-T receive array provided a 2-fold signal-to-noise ratio improvement over the 3-T array in the deep articular shoulder cartilage.

Conclusions: Shoulder imaging at 7-T is feasible with a custom transmit/receive array either in a single-channel transmit mode with a fixed RF shim or in a parallel transmit mode with a subject-specific RF shim.

FIGURE 1

Photographs of the 7-T shoulder transmit array (A), receive array (B), and both arrays (C).

FIGURE 2

“Unrolled” layout and electrical schematic of the 7-T shoulder transmit array (A) and receive array (B).

FIGURE 3

Simulated transverse 7-T B₁⁺ maps in a phantom for 3 transmit array configurations whose elements are highlighted in white. The elements are driven with a birdcage-like phase. Near–mirror symmetry is observed for the fully surrounding transmit array (center panel). The effect of B₁⁺ twisting emerges when the phantom is not fully surrounded by transmit coils (left and right panels). Given the polarity of the 7-T magnet at our facility, constructive interference occurs in the anterior region when the transmit array is on the left side of the phantom (arrowhead, right panel) and in the posterior region when the transmit array is on the right side of the phantom (arrowhead, left panel).

FIGURE 4

Flip angle magnitude for each transmit coil in the right shoulder of subject B generated using an 18-W hard pulse with 1-millisecond duration on each channel. The approximate location of the transmitting coils are highlighted in red and inactive transmit coils are outlined in white.

FIGURE 5

Anatomic reference image (A) and flip angle maps (B-F) in the right shoulder of subject A using various RF shim schemes. Approximate transmit coil locations are shown in white, along with corresponding RF shim weights. The ROI for the illustrated slice is highlighted in red in (A). The flip angle and nonuniformity listed at the bottom of each panel were calculated in ROIs in 7 consecutive slices, as described in the methods. The spatial reference point used to determine the geometric shim phases is denoted by an arrowhead in (A) and (B).

FIGURE 6

Anatomic reference image (A) and flip angle maps (B-F) in the left shoulder of subject A using various RF shim schemes.

FIGURE 7

Polar plots illustrate RF shim phase offsets for individual subjects and coils in the right (top group) and left (bottom group) shoulders. Tightly grouped colored lines indicate similarity between shim values despite anatomic variability among the subjects. The composite phase-coherent RF shim approach attempts to exploit this similarity (white lines).

FIGURE 8

Anatomic reference images and flip angle maps in the right and left shoulders of subject F at 7-T using the composite phase-coherent shim (middle row) and at 3-T using body coil (bottom row). All flip angle maps were generated using a 144-W rectangular excitation pulse with 1-millisecond duration, allowing direct comparison with Figures 5 and 6.

FIGURE 9

Three-plane in vivo SNR maps acquired using the developed 7-T 10-channel receive array (top row) and commercial 3 T 4-channel receive array (bottom row). In the deep cartilage (arrows in the left column), the 7-T 10-channel receive array provided a 2.1-fold SNR improvement over the 3-T 4-channel array.

FIGURE 10

Reciprocal g-factor (1/g) maps in a phantom for parallel imaging acceleration rates (R) of 2 to 4 in the anterior/posterior (top row), left/right (middle row), and head/foot directions (bottom row). The maximum g-factor is indicated in each panel.

FIGURE 11

The 7-T anatomic shoulder images. A to C, Reformatted 3D GRE images (0.7-mm isotropic) demonstrate good coverage, uniformity, fat suppression, and high SNR. D, proton density-weighted FSE image of the left shoulder depicts the sublabral foramen (closed white arrow) and anterior and posterior labra (closed black arrows). Good uniformity is observed. The articular cartilage can be visualized in D, E, and G (directly opposed black arrows). E, T2-weighted FSE image of the right shoulder reveals irregularity of the anterior labrum likely reflecting a partial tear (double white arrow). Lack of signal is observed in the peripheral muscles (black arrowhead), likely caused by high B₁⁺. F, T1-weighted TSE image of the right shoulder reveals normal muscle bulk for the rotator cuff in this healthy subject. G, T1-weighted TSE image of the left shoulder depicts the deltoid origin (double black arrow) and the supraspinatus tendon (white arrowhead) as well as the interface between the glenoid and humeral head articular cartilages (directly opposed black arrows). H, T2-weighted TSE image of the left shoulder shows hypertrophy of the anterior labrum (black closed arrow), posterior labral tear (white arrow), subchondral cyst formation (black open arrow), and thin posterior glenoid cartilage (black arrowhead), the latter of which may be related to posterior subluxation of the humeral head. I, T2-weighted TSE image of the left shoulder shows a partial-thickness articular surface tear of the posterior fibers of the supraspinatus (black arrowhead) and a subcortical cyst (double black arrow). J, Intermediate-weighted TSE image shows osteoarthritis manifested by cartilage thinning and osteophyte formation (ellipse) and a healed fracture deformity involving the greater tuberosity and surgical neck of the humerus (bracket).