Abstract
PURPOSE: Extend the universal pulse GRAPE formalism to pulses with a defined spectral response, and apply the concept to spatial selection. METHODS: We added Bloch simulations at several frequencies for each voxel to the pulse calculation to create universal spectrally-selective GRAPE pulses. With a superimposed constant gradient field spatial selection was achieved. The method was tested in slice- and slab-selective imaging experiments. RESULTS: Universal spatially-selective GRAPE pulses increased FA homogeneity and SNR. In 2D gradient echoes, the SNR could be increased by approximately 6% compared to CP pulses, and in a slab-selective TSE sequence, the SNR increased by 29% against k T -spokes pulses. Additionally, the slab-selective GRAPE pulse proved to be more robust against B0 deviations and is significantly shorter in comparison to k T -spokes pulses while maintaining a similar FA homogeneity. CONCLUSION: Spatially-selective universal GRAPE pulses exhibit superior performance compared to k T -spokes pulses. These short and robust pTx pulses hold potential for enhancing a wide range of imaging applications, thereby advancing 7T MRI technology closer to clinical use.