hide
Free keywords:
-
Abstract:
Purpose/Introduction: At ultra-high field (UHF, C7T), a simple increase of the length of a single-row human head transmit (Tx)-array cannot provide an adequate longitudinal coverage (1–4). Multi-row (C2) arrays and RF shimming have to be used instead (2, 3, 5). Tightfit surface loop transceiver (TxRx)-phased arrays (2, 4) improve Txefficiency
in comparison to Tx-only arrays (5), which are larger to fit
receive (Rx)-only arrays inside. Previously we demonstrated analytically (6) that at 9.4T both the magnetic and electric coupling between two loops can be compensated at the same time simply by overlapping, and excellent decoupling can be obtained for a single-row 8-element (1 9 8) human head Tx-array without additional decoupling strategies (7). In this work, we constructed a 9.4T human head 16-loop double-row (2 9 8) TxRx-array decoupled entirely by
overlapping. The array provides efficient transmission and the whole brain coverage.
Subjects and Methods: Loop size (10.5 cm x 10 cm) was first
evaluated analytically and then adjusted on a bench. The array (Fig. 1) measures 20 cm 9 23 cm (left–right 9 anterior–posterior), and 17.5 cm in length. Overlapping provides very good decoupling (Fig. 1). To decrease radiation losses the array is shielded. We compared the performance of the TxRx-array with a larger 16-element
Tx-only/30-element Rx-only(ToRo) surface loop phased array
(28 cm–diameter, 19 cm-length) described previously (5). Electromagnetic(EM) simulations of the B1 + and the local SAR were performed using CST Studio Suite 2015 (CST, Darmstadt, Germany). Three voxel models were used, i.e. a head/shoulder (HS) phantom (5), and two multi-tissue models, ‘‘Duke’’ and ‘‘Ella’’. Experimental B1 + maps were obtained using the AFI sequence (8). Data were acquired on a Siemens Magnetom 9.4T human imaging system.
Results: After EM modeling, we conducted in vivo experiments
(Fig. 2). The tight-fit TxRx-array provides *45 improvement in B1 +-efficiency (B1 +/HP) compared to the ToRo-array, which agrees well with EM simulations. Circular polarizition (CP) produces an inhomogeneous B1 + at UHF (Fig. 2A, B). Introduction of a 658-phase shift, ß, between the rows (Fig. 2C) improves homogeneity (9). Invivo images (Fig. 3A) demonstrate whole-brain coverage. Figures 3B, C show SNR maps obtained in vivo using both arrays. While the ToRo-array has a better peripheral SNR, SNRs in the center are similar. We also evaluated the dependence of the maximum local SAR10g on b. At b * 658 maximal SAR10g decreased by 13 as compared to the CP Mode.
Discussion/Conclusion: We constructed a 9.4T (400 MHz) 2 9 8
head TxRx-array based on the results of the analytical modeling. We demonstrated that simply by overlapping good decoupling can be obtained without additional decoupling strategies. This provides a recipe for a simple, robust, very Tx-efficient design suitable for parallel transmission and whole brain imaging at UHF.
