Benchmarking Gradient Impulse Response Function (GIRF) Acquisition Strategies with Open-Source Measurement Software.

Motivation: Gradient impulse response functions (GIRF) can correct MR imaging gradient-based errors that plague various techniques. Methods exist for acquiring and processing GIRFs, but there are few tools for comparisons. Goal(s): To demonstrate an open-source GIRF acquisition, processing, and benchmark package (github.com/mloecher/GIRFbench), and describe initial testing strategies. Approach: A range of test waveforms were designed in Pulseq to estimate the GIRF using both a field camera and thin-slice MRI-based methods. We compared the performance of these different GIRF estimation strategies, including post-processing filter choices. Results: Estimated GIRFs reduced PC-MRI background velocity errors to ≤0.4% of VENC and corrected spiral trajectories to within ∆k=0.31 m-1. Impact: We developed an open-source package (github.com/mloecher/GIRFbench) to acquire, process, and benchmark GIRF measurement strategies to better standardize and understand these techniques. Estimated GIRFs reduced PC-MRI background velocity errors to ≤0.4% of VENC and corrected spiral trajectories to within ∆k=0.31 m-1.