AZTEK: Adaptive zero TE k‐space trajectories

Abstract

Purpose Because of short signal lifetimes and respiratory motion, 3D lung MRI is still challenging today. Zero‐TE (ZTE) pulse sequences offer promising solutions as they overcome the issue of short T2∗. Nevertheless, as they rely on continuous readout gradients, the trajectories they follow in k‐space are not adapted to retrospective gating and inferred motion correction. Theory and methods We propose AZTEK (adaptive ZTE k‐space trajectories), a set of 3D radial trajectories featuring three tuning parameters, to adapt the acquisition to any moving organ while keeping seamless transitions between consecutive spokes. Standard ZTE and AZTEK trajectories were compared for static and moving phantom acquisitions as well as for human thoracic imaging performed on 3 volunteers (1 healthy and 2 patients with lung cancer). Results For the static phantom, we observe comparable image qualities with standard and AZTEK trajectories. For the moving phantom, spatially coherent undersampling artifacts observed on gated images with the standard trajectory are alleviated with AZTEK. The same improvement in image quality is obtained in human, so details are more delineated in the lung with the use of the adaptive trajectory. Conclusion The AZTEK technique opens the possibility for 3D dynamic ZTE lung imaging with retrospective gating. It enables us to uniformly sample the k‐space for any arbitrary respiratory motion gate, while preserving static image quality, improving dynamic image quality and guaranteeing continuous readout gradient transitions between spokes, which makes it appropriate to ZTE.

Motion corrected silent ZTE neuroimaging

“…Several methods exist to produce interleaved spiral trajectories 27,32,33 . We adopted the 3D spiral phyllotaxis method because of its pseudo‐random sampling and elegant mathematical description 27 .…”

“…Several methods exist to produce interleaved spiral trajectories. 27,32,33 We adopted the 3D spiral phyllotaxis method because of its pseudo-random sampling and elegant mathematical description. 27 In the original phyllotaxis trajectory design, the full set of spokes is first obtained via constant increments of (i) the azimuthal angle by the golden angle (𝜙 G ≈ 137.5 • ) and (ii) the z-coordinate such that the first and last spoke align along the north and south pole of the k-space sphere.…”