Practical Implementation of UTE Imaging

Abstract

“Ultrashort TE” (UTE) has gained attention in recent years because of its ability to visualize tissues with very short T2 that cannot normally be seen with MRI. In principle, the method for implementing 2-D UTE measurements is quite simple, requiring the use of selective excitation by half RF pulses, rapid readout of a half echo, and summation of pairs of excitations with opposite gradient polarity. In practice, imperfections in a real MR imager, such as eddy currents or distorted gradient waveforms, disturb the summation process (linear superposition) causing poor slice selection as well as shading and other artifacts. The aim of this work was to understand the various sources of error and to find reliable methods to correct them. It is shown that the slice-selection process and the readout process can be analyzed separately. A proposed system correction compensates for distortions in read gradient waveforms for each axis and for a range of gradient amplitudes. A proposed correction for excitation errors maintains proper timing of the two partial excitations to be added for any oblique plane within the magnet. A proposed correction for final readout timing errors interacts with the gridding reconstruction process to find correct locations for the samples in k-space. Corrections of this kind can lead to reliable routine clinical use of UTE with good image quality. Keywords: ultrashort TE (UTE); half RF; half echo; radial sampling; readout correction; gridding correction; iterative correction; variable rate selective excitation (VERSE)