Sparsity-enforced slice-selective MRI RF excitation pulse design

Area of Science:

• Magnetic Resonance Imaging (MRI)
• Radio-Frequency (RF) Pulse Design
• Sparse Approximation Theory

Background:

• Designing effective MRI pulses is crucial for accurate imaging.
• Conventional methods struggle with B(1) inhomogeneity and complex excitation patterns.
• Optimizing pulse design can lead to faster and more precise imaging.

Purpose of the Study:

• Introduce a novel algorithm for fast, slice-selective, spatially-tailored MRI excitation pulses.
• Utilize sparse approximation and second-order cone optimization for pulse design.
• Mitigate B(1) inhomogeneity and achieve user-defined magnetization patterns.

Main Methods:

• Developed a novel algorithm based on sparse approximation theory.
• Employed second-order cone optimization to place and modulate RF pulse segments (spokes) in k-space.
• Enforced sparsity on the number of spokes while optimizing their placement and modulation.
• Tested pulses in water and oil phantoms at 7 T and 3 T, respectively.

Main Results:

• Sparsity-enforced pulses outperformed conventional Fourier-based pulses in phantom experiments.
• A 5.7-ms, 15-spoke pulse achieved 1.28x lower RMSE for uniform magnetization under B(1) inhomogeneity compared to conventional pulses.
• Conventional methods required longer (7.8 ms) and more spokes (29) for similar uniformity.

Conclusions:

• The novel sparsity-enforced algorithm enables faster and more accurate MRI pulse design.
• This method effectively mitigates B(1) inhomogeneity and produces highly-structured excitations.
• Outperforms conventional techniques in achieving target magnetization with fewer, shorter RF pulses.