In current clinical cardiac MR (CMR) integrated body RF coils are commonly used for ¹H excitation, which deposit the RF energy/SAR in a large volume of the patient/subject. Progress in ultrahigh field MR (B₀≥7T), where body RF coils are not provided, demonstrated that local transceiver (TX/RX) RF coil arrays are suitable for RF excitation of deep lying regions such as the heart^(1,2). Recognizing this opportunity, this work demonstrates the feasibility of CMR at 3T with a local four-channel TX/RX RF coil array^(3,4). This setup is benchmarked against the clinical standard of a body RF coil for excitation in conjunction with a four-channel RX-only RF coil. Our assessment includes electromagnetic field (EMF) simulations to detail SAR and B₁⁺-homogeneity and -efficiency for the local and body RF coil configuration and their in vivo performance for high spatial resolution 2D SSFP-CINE imaging of the heart.

12 volunteers underwent CMR at a 3T whole-body MR system (MAGNETOM Verio, Siemens Healthcare, Erlangen, Germany). A four-channel transceiver RF coil (4TX/4RX) was constructed with 4 rectangular loop elements (H-F=180mm, L-R=120mm) (Fig 1). Basic B₁⁺-phase shimming was facilitated by phase-shifting cables. The TX phase setting Φ was derived from EMF simulations (CST AG, Darmstadt, Germany) using voxel model Duke. Φ minimizes

$$f(\phi)=\frac{std(B_1^+(\phi))}{mean(B_1^+(\phi))}-\frac{MOS(B_1^+(\phi))}{SOM(B_1^+(\phi))}\frac{1}{\sqrt{\max(SAR_{10g}(\phi))}}$$

so simultaneously optimizes B₁⁺-homogeneity (1st term), -efficiency (2nd term) and local SAR_10g (3rd term). |B₁⁺| denotes the magnetization within Duke’s heart, MOS/SOM is the magnitude of sum/sum of magnitude. Rapid SAR_10g computations were conducted with pre-calculated 4x4-SAR-matrices (SimOpTx, Vienna, Austria). For all simulations maximum SAR_10g was calculated and used to determine the maximum RF input power according to IEC guidelines (in normal operating mode)⁽⁵⁾. The 4TX/4RX RF coil was benchmarked against the built-in body RF coil (BC) together with a home-built four-channel receive-only RF coil (4RX), resembling the 4TX/4RX coil geometry. Protocol parameters of invivo 2D Bloch-Siegert-B₁⁺-mapping: spatial resolution=5.3x5.3x6mm³, 4.5ms Fermi pulse, off-center frequency: 4kHz, TE/TR=7.3/80ms, TA=15s. Based on the B₁⁺-maps flip angle maps were calculated. Protocol parameters of invivo 2D SSFP CINE technique: spatial resolution=1.8x1.8x6mm³, 30 cardiac phases, TA=15s (within one breath hold), TE/TR=1.4/3.2ms, FA maximal to reach SAR limit. For the body RF coil an additional SSFP protocol with a reduced flip angle FA’ was acquired, which assumes local instead of whole-body SAR limits:

$$FA'=\sqrt{\frac{local SAR limit}{whole-bodySARlimit}}*\sqrt{\frac{whole-body SAR_{Simulation}}{\max(localSAR_{Simulation})}}*FA\approx0.7*FA$$

To increase FA_ROI TR_SSFP was changed by varying the receiver bandwidth BW_RX, while keeping all other imaging parameters constant.

The SAR simulation of the body coil showed that the local SAR_10g limit is by a factor of 40% more restrictive than the whole-body SAR limit (Fig 2). Therefore the local SAR limit was additionally used for the body coil. The transmit efficiency $$$B_1^+@SARlimit=B_1^+*\sqrt{\frac{SARlimit}{SAR_{Simulation}}}\sim FA$$$ was chosen to compare the transmission regimes: within Duke’s heart B₁⁺@SAR-limit=130/92µT for BC@whole-body/local SAR limit, and 97µT for 4TX/4RX@local SAR limit (Fig 3). The simulated transmit homogeneity std(B₁⁺)/mean(B₁⁺) was 15% for BC/4RX and 30% for 4TX/4RX. The quality of the in vivo images of all transmission regimes was clinically acceptable and not affected by signal voids. The transmit efficiency advantage of 4TX/4RX manifests in a higher mean FA_ROI (Fig 4 top row), while the BC is more B₁⁺-homogenous. For an apical SAX this efficiency of the 4TX/4RX setup (TR_min=3.8ms) reduced TR_min by 29%/54% compared to the body RF coil transmission at whole-body/local SAR limit (TR_min=4.7/8.3ms). The TR_min shortening of 4TX/4RX helped to reduce SSFP banding artifacts and relaxed the constraints on the fidelity of volume-selective B₀-shimming.Although whole-body SAR limits are used for the body RF coil in clinical cardiac MR, local SAR_10g limits are more restrictive. When applying the maximum power to reach the whole-body SAR limit with the body RF coil the local SAR_10g limit is already exceeded by 90%, which agrees with ^(6,7). The short distance to the human chest renders the local transmit RF coil more efficient than the body RF coil in terms of transmit efficiency $$$ B_1^+/\sqrt{P_{in}} $$$ as well as $$$B_1^+/\sqrt{localSAR_{10g}}$$$. High transmission efficiency is clinically relevant for CMR, where cardiac and respiratory motion makes rapid imaging techniques such as SSFP mandatory. Moreover, body RF coil transmission induces RF power deposition in a large volume including body regions far away from the ROI, which constitute an RF-heating related contraindication for patients with implants⁽⁸⁾. Local transmit RF coils restrict the volume of power deposition and hence permit inclusion of these patients into CMR examinations. To conclude, pursuing local TX/RX RF coil arrays in CMR is a conceptually appealing alternative to body RF coil transmission and has high clinical potential and implications for future’s MRI.