Open-Source, PNS-Constrained, and Optimized Gradient Waveform Design for Brain Diffusion Tensor Imaging

Motivation: Gradient optimization approaches are an excellent strategy for DTI to manage competing constraints, particularly PNS on ultra-high-performance systems; however, vendor-specific implementations limit accessibility. Goal(s): To integrate the open-source Gradient Optimization Toolbox (GrOpt) in Pulseq for brain DTI. Approach: We implemented a 2x-accelerated, PNS constrained, single-shot EPI sequence in PyPulseq to include both GrOpt and conventional trapezoidal diffusion waveforms. The protocol was evaluated on commodity hardware and ultra-high-performance hardware. Results: GrOpt-enabled sequences with PNS control demonstrated significant TE reductions compared to conventional waveforms. DTI metrics were validated across volunteers, with GrOpt consistently achieving higher SNR on both standard and ultra-high-performance MRI systems. Impact: Integrating GrOpt into Pulseq enables efficient, SNR-enhanced diffusion encoding on ultra-high-performance MRI systems while managing peripheral nerve stimulation (PNS). This open-source solution makes advanced pulse sequence design accessible, thereby enabling DTI protocol consistency and utility across research settings.