$B_1^{+}$ inhomogeneity mitigation for diffusion weighted MRI at 7T using TR-FOCI pulses

Purpose: The purpose of this study is to improve the image quality of diffusion-weighted images obtained with a single RF transmit channel 7 T MRI setup using time-resampled frequency-offset corrected inversion (TR-FOCI) pulses to refocus the spins in a twice-refocused spin-echo readout scheme.

Methods: We replaced the conventional Shinnar-Le Roux-pulses in the twice refocused diffusion sequence with TR-FOCI pulses. The slice profiles were evaluated in simulation and experimentally in phantoms. The image quality was evaluated in vivo comparing the Shinnar-Le Roux and TR-FOCI implementation using a b value of 0 and of 1000 s/mm².

Results: The b0 and diffusion-weighted images acquired using the modified sequence improved the image quality across the whole brain. A region of interest-based analysis showed an SNR increase of 113% and 66% for the nondiffusion-weighted (b0) and the diffusion-weighted (b = 1000 s/mm²) images in the temporal lobes, respectively. Investigation of all slices showed that the adiabatic pulses mitigated $B_1^{+}$ inhomogeneity globally using a conventional single-channel transmission setup.

Conclusion: The TR-FOCI pulse can be used in a twice-refocused spin-echo diffusion pulse sequence to mitigate the impact of $B_1^{+}$ inhomogeneity on the signal intensity across the brain at 7 T. However, further work is needed to address SAR limitations.