Fast, simple gradient delay estimation for spiral MRI

Abstract

Timing delays between data acquisition and gradient transmission result in image degradation. This is especially true in spiral MRI, where delays can alter data in a nonuniform manner, generating significant artifact in the reconstructed data. The many methods that exist to mitigate these delays or measure the k-space coordinates require long measurement times, complicated analysis, specialized phantoms or hardware, or significant changes to the sequence of interest. A fast and simple method is proposed to measure delays on each gradient channel. It requires only minimal modification to an existing spiral sequence and can be used to measure independent delays on three gradient channels and any scan subject within six sequence repetition times. Accurate reconstruction of MRI data requires knowledge of the k-space sample locations. Deviations from the theoretical gradient waveforms can be caused by eddy currents, gradient amplifier nonlinearities, and other system imperfections. These deviations take the form of gradient timing delays and amplitude distortions that result in altered sample locations in k-space, producing artifacts in the reconstructed images. In spiral trajectories, these artifacts can be mistaken for aliasing, loss of resolution, poor signalto noise-ratio, or off-resonance blurring, thus making the source of the artifacts difficult to discern.Modern MRI scanners mitigate eddy currents through the use of gradient coil shielding and gradient waveform preemphasis. Sequences that place large gradient slew-rate and amplitude demands on the hardware can induce eddy currents, despite gradient coil shielding. Gradient waveform pre-emphasis provides an additional level of eddy current compensation but can only be optimized for a limited range of sequences. A number of techniques have been employed to compensate for the remaining deviations from the expected k-space coordinates.Several strategies seek to measure the actual gradient waveform for correct reconstruction or improved gradient pre-emphasis. Spielman and Pauly (1) directly measured the current on each gradient channel. Errors from the gradient amplifier were directly measured, but eddy currents were assumed to be negligible. Onodera et al. (2) proposed

Single‐shot spiral imaging at 7 T

“…The third model ingredient, field dynamics, can be determined by a range of methods. Spiral trajectories have been mapped previously using additional reference scans on a phantom or the subject itself . More recently they have also been predicted based on gradient impulse response functions .…”

Abdominal 4D Flow MR Imaging in a Breath Hold: Combination of Spiral Sampling and Dynamic Compressed Sensing for Highly Accelerated Acquisition

“…There are multiple ways to improve the image quality. First, spiral reconstruction may benefit from correction techniques (21, 22). Second, compressed sensing reconstruction may use extra regularization terms promoting angiogram (23, 24) or phase characteristic (25) properties.…”

Whole-brain background-suppressed pCASL MRI with 1D-accelerated 3D RARE Stack-Of-Spirals readout

“…At the beginning of the sequence, gradients were rapidly played with alternating polarity to correct for their delay in the spiral trajectories [19], followed by two preparation TRs, to allow the signal to reach the steady state. A non-accelerated readout was played during the preparation TRs, in order to obtain a fully sampled k-space dataset, used for calibration of the parallel imaging reconstruction kernel, needed to reconstruct the skipped k z partitions in the accelerated images.…”

“…The spiral trajectories were corrected prior to reconstruction by estimating the gradient time delays [19]. Spiral k-space data were then regridded to Cartesian coordinates using the standard regridding method with a Kaiser-Bessel window kernel [20], and a Voronoi diagram to correct non-uniform sampling density.…”