# Phantom‐based gradient waveform measurements with compensated variable‐prephasing: Description and application to EPI at 7 T

## TL;DR

The paper introduces a new method for measuring gradient waveforms in MRI that improves accuracy and helps correct image distortions.

## Contribution

The novel 'compensated variable-prephasing' method enhances gradient waveform measurements by accounting for all gradients involved.

## Key findings

- CVP provides more reliable gradient measurements than VP and FCVP in the tested setup.
- All three methods resulted in high-quality EPI images after trajectory corrections.
- VP measurements are affected by lingering oscillations not present in CVP and FCVP.

## Abstract

Introducing compensated variable‐prephasing (CVP), a phantom‐based method for gradient waveform measurements. The technique is based on the variable‐prephasing (VP) method, but takes into account the effects of all gradients involved in the measurement.

We conducted measurements of a trapezoidal test gradient and of an EPI readout gradient train with three approaches: VP, CVP, and fully compensated variable‐prephasing (FCVP). We compared them to one another and to predictions based on the gradient system transfer function. Furthermore, we used the measured and predicted EPI gradients for trajectory corrections in phantom images on a 7 T scanner.

The VP gradient measurements are confounded by lingering oscillations of the prephasing gradients, which are compensated in the CVP and FCVP measurements. FCVP is vulnerable to a sign asymmetry in the gradient chain. However, the trajectories determined by all three methods resulted in comparably high EPI image quality.

We present a new approach allowing for phantom‐based gradient waveform measurements with high precision, which can be useful for trajectory corrections in non‐Cartesian or single‐shot imaging techniques. In our experimental setup, the proposed “compensated variable‐prephasing” method provided the most reliable gradient measurements of the different techniques we compared.