Slice-selective Zero Echo Time imaging of ultra-short T2 tissues based on spin-locking

TL;DR

This paper introduces two novel spin-locking integrated ZTE sequences, DiSLoP and PreSLoP, enabling slice-selective imaging of ultra-short T2 tissues with improved efficiency and SNR, expanding ZTE capabilities for challenging tissues.

Contribution

The paper presents new slice selection methods for ZTE imaging that effectively target ultra-short T2 tissues using spin-locking, a significant advancement over existing techniques.

Findings

Achieved control over slice profiles and positions in 2D imaging.

Demonstrated imaging of tissues with T2 as low as 275 microseconds.

PreSLoP reduces scan time by 2-5 times compared to standard 3D ZTE.

Abstract

Purpose: To expand the capabilities of Zero Echo Time (ZTE) pulse sequences with a slice selection method suitable for the shortest-lived tissues in the body. Methods: We introduce two new sequences that integrate spin-locking pulses into standard ZTE imaging to achieve slice selection: one for moderately short $T_2$ (DiSLoP), the other for ultra-short $T_2$ samples (PreSLoP). These methods exploit the slower signal decay (at $T_{1\rho}\gg T_2$) to retain the magnetization in the slices during the selection process, which is otherwise comparable to or even much longer than $T_2$. Results: We demonstrate control over the slice profiles and positions for 2D imaging. We measure magnetization decay times during spin-locking ($T_{1\rho}$) as a function of pulse amplitude, showing significant lifetime enhancement for amplitudes as low as 10 uT. We show imaging of slice-selected samples with $T_2$ characteristic times in the range of single milliseconds with DiSLoP and PreSLoP, and with the latter for sub-millisecond $T_2$ tissues. As compared to standard 3D ZTE sequences, PreSLoP achieves the same signal-to-noise ratio (SNR) in 2-5 times shorter scan times, and we argue that this is due to the filling scheme of the finite gap at the center of $k$-space unavoidable with ZTE sequences. Finally, we discuss a combination of DiSLoP with a dynamical decoupling sequence to avoid this central gap, leading to further scan time accelerations. Conclusions: The proposed sequences are capable of slice-selected 2D imaging of tissues with $T_2$ as low as 275 us with good SNR within clinically acceptable scan times.