Design of spectral‐spatial phase prewinding pulses and their use in small‐tip fast recovery steady‐state imaging

    loading  Checking for direct PDF access through Ovid

Excerpt

Inhomogeneity of the B0 field can cause spatially varying phase and signal loss in MRI. Spin‐echo sequences can realign static off‐resonance effects, but require a second refocusing RF pulse and may be unsuitable for short TR imaging or applications that are specific absorption rate (SAR) limited 1. Assländer et al. introduced spectral prewinding pulses to compensate for spin dephasing in rapid gradient echo sequences (GRE), effectively combining the speed of GRE with the off‐resonance robustness of spin‐echo 2, for a restricted off‐resonance bandwidth.
Recently, spectrally selective phase prewinding pulses were incorporated into the small‐tip fast recovery (STFR) steady‐state sequence 3. The STFR sequence employs both a traditional “tip‐down” pulse and then a “tip‐up” pulse that quickly returns the magnetization to the longitudinal axis after free precession 4. STFR has mixed T2/T1 contrast and provides nearly equal signal level and tissue contrast in the human brain as balanced steady‐state free precession 5. One challenge with balanced steady‐state free precession is that banding artifacts may occur 6 that can only be removed with multiple acquisitions and phase cycling 7. With appropriately designed “tip‐up” pulses, STFR can potentially eliminate the banding artifact in a single acquisition 4.
However, purely spectral prewinding pulses can refocus a limited range of frequencies 2. This article proposes a spectral‐spatial pulse that increases the effective prewinding bandwidth and mitigates the limitations of a purely spectral design. The key idea of this paper is to vary spatially the pulse design criterion based on voxel‐by‐voxel bulk off‐resonance. The resulting pulse accommodates a wider effective bandwidth, improving signal recovery in inhomogeneous objects. Our proposed application for spectral‐spatial STFR is brain imaging, where inhomogeneity in the B0 field is generally smoothly varying with a wide bandwidth.
In the following section, we review spectral prewinding pulse design and introduce the extension to spectral‐spatial prewinding pulses. We then formulate our constrained pulse design optimization problem under the small‐tip angle (STA) approximation 8. Next, we outline pulse design validation experiments in a phantom and in vivo human brain. We quantify performance with subsequently defined metrics. Finally, we present simulation and experimental results for the phantom and in vivo pulse designs.

Related Topics

    loading  Loading Related Articles