Purpose: Arterial-spin-labeling (ASL) magnetic resonance imaging (MRI) is intrinsically a low signal-to-noise ratio (SNR) technique. This study aims to enhance its sensitivity by quantifying physiological noise attributed to cardiac pulsation and devising an improved sequence, cardiac-triggered ASL, to minimize this noise. Methods: A total of 16 healthy subjects were studied on a 3 Tesla MRI system. The influence of cardiac pulsation on pseudo-continuous ASL (pCASL) signal was first investigated by performing a regular pCASL sequence while the cardiac phase of the image acquisition was recorded by a pulse oximeter. We then designed a new sequence, cardiac-triggered pCASL, to align the cardiac phases of the control and labeled scans. The performance of the new sequence was evaluated in the context of single-shot 3D gradient-and-spin-echo acquisition, multishot 3D spiral acquisition, and hypercapnia-induced cerebral blood flow (CBF) changes. Results: In regular pCASL, the signal intensities of both control and labeled images were strongly modulated by the cardiac phase. In single-shot acquisitions, this results in signal instability in regions near large vessels. In segmented acquisitions, it results in ghosting artifacts in the image and, furthermore, the signal fluctuation is smeared along the segmentation direction to affect more brain regions. Cardiac-triggered pCASL enhanced the temporal SNR by 94% and 28% in single-shot and segmented 3D acquisition, respectively. When applied to detect CBF changes, the triggered sequence revealed a greater statistical power in terms of both the number of significant voxels and t-score histograms. Conclusion: Cardiac-triggered pCASL represents a potential scheme to enhance the reliability of ASL signal. Magn Reson Med 80:969–975, 2018.
- arterial spin labeling
- cardiac phase
ASJC Scopus subject areas
- Radiology Nuclear Medicine and imaging