Yunhong Shu, Presenter: Nothing to Disclose

Andrew M. Elliott PhD, Abstract Co-Author: Nothing to Disclose

Stephen J. Riederer PhD, Abstract Co-Author: Nothing to Disclose

Matthew A. Bernstein PhD, Abstract Co-Author: Nothing to Disclose

A shell k-space trajectory was implemented for 3D imaging to enable a novel motion correction method. A self-navigating property of the shell trajectory enables the tracking and correction of arbitrary rigid body motion.

The shell trajectory covers a spherical 3D k-space volume with a set of concentric spherical shells of increasing radii. Each shell is sampled by helical spiral interleaves that sweep between the two poles. During the acquisition, a short time interval corresponds to the coverage of a spherical shell. If we only use that shell to reconstruct a 3D tracking image, a point marker shows a 3D bull’s-eye pattern. The true location of a point marker corresponds to the center of the pattern. Three point markers were affixed to the imaged subject, enabling determination of rigid body translation and/or rotation during each time interval. With the knowledge of the 3D coordinates of the markers, we could track the motion history of the subject to approximately 1mm. Phase corrections and rotations of k-space are used to retrospectively correct the artifacts caused by translations and rotations, respectively.

64 concentric shells constructed from a total of 1476 interleaves yielded a sampling matrix of 128x128x128 with a 240mm FOV. At ±62.5 KHz BW, the readout time for each interleaf was 8ms. Total acquisition time was 22sec for TR=15ms. The sampled data were reconstructed using gridding algorithm. A computer-based simulation was done to test the trajectory’s performance under off-resonance effects. The PSF showed that the new shell trajectory performed well both under ideal conditions and in the presence of off-resonance. The attenuation of the PSF was 17% under a 60Hz constant off-resonance frequency in the z direction. A head scan was acquired on a volunteer who nodded his head during the acquisition. The motion correction results showed that the method removed a substantial amount of artifacts caused by the motion. The motion-corrected images yielded similar image quality to the reference images without motion.

The self-navigating shell trajectory is useful for arbitrary rigid body motion correction in 3D acquisitions.

Shu, Y, Elliott, A, Riederer, S, Bernstein, M, 3D Ringlet: Use of a Concentric Spherical Shell K-Space Trajectory for Rigid Body Motion Correction in 3D Imaging.  Radiological Society of North America 2004 Scientific Assembly and Annual Meeting, November 28 - December 3, 2004 ,Chicago IL. http://archive.rsna.org/2004/4404888.html