Meeting Banner
Abstract #1402

Compressed Sensing 3D Ultrashort Echo Time (COMPUTE) Imaging

MAGNA25Cheng Li1, Jeremy Magland1, Felix Wehrli1

1Radiology, University of Pennsylvania, Philadelphia, PA, United States

Ultra-short echo time (UTE) imaging is a technique for directly imaging tissues with sub-millisecond T2 values. However, 2D UTE imaging is inherently time-inefficient because of half-pulse excitation and radial center-out sampling. In this work we describe 3D Compressed Sensing UTE (COMPUTE) imaging with a hybrid-radial encoding strategy to improve efficiency. Phantom and in vivo results demonstrate the feasibility of COMPUTE technique with an acceleration factor of 10. In work in progress, long-T2 suppression modules are being incorporated into COMPUTE, allowing spatial sparsity of the soft-tissue suppressed images to be exploited to achieve even higher acceleration factors.

Keywords

acceleration accomplished according accuracy achieve achieved achieving allowing amplitude analogous anisotropic anticipate apparent artifacts axial bandwidth channel chosen coil combined compensation compressed compute computed conjugate considering constraint consuming coronal correspond coverage currently custom dashed decay degradation demonstrating denotes density designed developed dimension directly distance duration effective efficiency employed encoding equiangular evaluate evolved excitation fast feasibility finite full fully function gradient gradients half hard head highest hybrid immune impractically improve in vivo incorporated inefficient inherently knee long loss male maps microseconds minimize minimizing minutes modules multiplying need norm obviates onto operator opposite optimization pattern perceptible phantom plotted polarities press probability profiles progress projection projections promising pulse pulses quality radial radiology ramp random rather readout reasons reconstructed reconstruction reconstructions reduced regularization sampled sampling selection selectivity sensing sets short significantly simulation since slice smoothing solving space sparsity spatial spherical still straightforward strategy suppressed target thereby tibia trajectory transform ultra uniform variably variation views volunteer year zero