Improving Zero TE imaging with RF encoding
Publication Date: 2015-Mar-06
The IP.com Prior Art Database
Title: Improving Zero TE imaging with RF encoding
The approach relates to a magnetic resonance imaging method which involves a zero-echo time (ZTE) acquisition sequence. The acquisition employs radial k-space scanning in which subsequently scanned radial lines have only a small angular offset in k-space.
According to the approach, magnetic resonance signals acquired from the centre region of k-space are spatially encoded by way of spatial RF encoding based on B1-field gradients.
The approach achieves spatial encoding of the magnetic resonance signals acquired at or near the centre of k-space while avoiding problems due to finite duration of RF excitation and finite duration of T/R switching.
Zero echo time (ZTE) imaging offers the possibility to image tissue with short transverse relaxation (bones, teeth, etc.), at significantly reduced acoustic noise. However, ZTE is hampered by reliably sampling central k-space due to finite duration of RF excitation and finite duration of TX/RX switching.
This approach suggests a further and new alternative to sample inner k-space data in ZTE by means of spatial RF encoding. RF encoding is based on B1 gradients, realized via suitably designed RF pulses and an appropriate superposition of local TX elements, producing predefined spatial distributions of phase and amplitude.
Sampling of inner ZTE k-space can be done in two different ways: (1) sampling single points near k-space center by RF encoding without using any B0 gradients, or (2) combining B0 and B1 gradients, i.e., shifting the B0 encoded k-space lines towards k-space center by corresponding B1 gradients.
Zero echo time (ZTE) imaging offers the possibility to image tissue with short transverse relaxation, achieved by 3D radial center-out trajectories and hard-pulse RF excitation applied already in presence of the read-out gradient. The gradient is not ramped to zero between profiles, actually it is kept fixed of constant amplitude, but it is just modified step-wise, to change direction, leading to practically silent MRI.
As in standard MRI, spatial encoding of ZTE is based on B0 gradients, producing spatially linear distributions of phase and frequency. Alternatively, spatial encoding can be based on B1 gradients, realized via suitably designed RF pulses of local TX elements (“RF encoding”), producing predefined (not necessarily linear) spatial distributions of phase and amplitude.
ZTE is hampered by reliably sampling data in the k-space centre. The dead time due to finite duration of RF excitation and finite duration of TX/RX switching causes a start of sampling several voxels apart from k-space center. This means ZTE has a blind spot in the centre of k-space and different measures have been proposed to solve this problem.
One approach, the algebraic reconstruction, is trying to estimate the missing sample data from the measured ones including appropriate signal oversampling. The approach breaks down if t...