Optimization of Oscillating Diffusion Encoding Gradient Reduces Mechanical Vibration on Ultra-high Gradient System

Motivation: OGSE (Oscillating Gradient Spin Echo) is a powerful tool for time-dependent diffusion MRI to detect cell structures at the micron scale. It requires strong gradient to provide b-value at a certain oscillating frequency, which will introduce significant mechanical vibration in the gradient coils. Goal(s): Our study aims to optimize the trapezoidal oscillating gradient to mitigate the mechanical vibration. Approach: A convex optimization problem is designed to reduce vibration by minimizing vibration level predicted by the system frequency response function, while ensuring the b-value and OG encoding spectrum through multiple constraints. Results: Optimized OG waveform can achieve vibration reduction up to 3dB on RMS. Impact: OG caused mechanical vibration poses significant challenge to MRI hardware, especially in high-performance gradient system. Our vibration optimized OG waveforms can improve subject's comfort and have potential to enable more precise parameter fitting, enhancing the accuracy of microstructural imaging.