Schwarzbauer Christian, Davis Matt H, Rodd Jennifer M, Johnsrude Ingrid
MRC Cognition and Brain Sciences Unit, Cambridge, UK.
Neuroimage. 2006 Feb 1;29(3):774-82. doi: 10.1016/j.neuroimage.2005.08.025. Epub 2005 Oct 14.
The acoustic scanner noise that is generated by rapid gradient switching in echo planar imaging (EPI) is an important confounding factor in auditory fMRI. "Sparse imaging" designs overcome the influence of scanner noise on stimulus presentation by acquiring single brain volumes following a silent stimulus presentation period. However, conventional sparse imaging requires assumptions about the time-to-peak of the evoked hemodynamic response and reduces the amount of EPI data which can be acquired and hence statistical power. In this article, we describe an "interleaved silent steady state" (ISSS) sampling scheme in which we rapidly acquire a set of EPI volumes following each silent stimulus presentation period. We avoid T1-related signal decay during the acquisition of the EPI volumes by maintaining the steady state longitudinal magnetization with a train of silent slice-selective excitation pulses during the silent period, ensuring that signal contrast is constant across successive scans. A validation study comparing ISSS to conventional sparse imaging demonstrates that ISSS imaging provides time course information that is absent in conventional sparse imaging data. The ISSS sequence has a temporal resolution like event-related (ER) imaging within a single trial (unlike conventional sparse imaging, where ER-like temporal resolution can only be achieved by compiling data across many jittered trials of the same stimulus type). This temporal resolution within trials makes ISSS particularly suitable for experiments in which a) scanner noise would interfere with the perception and processing of the stimulus; b) stimuli are several seconds in duration, and activation is expected to evolve and change as the stimulus unfolds; and c) it is impractical to present a single stimulus more than once (for example, repetition priming or familiarity effects would be expected).
回波平面成像（EPI）中快速梯度切换所产生的声学扫描仪噪声是听觉功能磁共振成像中的一个重要混杂因素。“稀疏成像”设计通过在无声刺激呈现期之后采集单个脑容积来克服扫描仪噪声对刺激呈现的影响。然而，传统的稀疏成像需要对诱发的血液动力学反应的峰值时间做出假设，并减少了可采集的EPI数据量，从而降低了统计功效。在本文中，我们描述了一种“交错无声稳态”（ISSS）采样方案，即在每个无声刺激呈现期之后快速采集一组EPI容积。我们通过在无声期用一系列无声切片选择激发脉冲维持稳态纵向磁化，避免在EPI容积采集期间出现与T1相关的信号衰减，确保在连续扫描中信号对比度保持恒定。一项将ISSS与传统稀疏成像进行比较的验证研究表明，ISSS成像提供了传统稀疏成像数据中所没有的时间进程信息。ISSS序列在单次试验中具有类似事件相关（ER）成像的时间分辨率（与传统稀疏成像不同，传统稀疏成像只有通过汇总同一刺激类型的许多抖动试验的数据才能实现类似ER的时间分辨率）。试验内的这种时间分辨率使得ISSS特别适用于以下实验：a）扫描仪噪声会干扰刺激的感知和处理；b）刺激持续数秒，并且预期随着刺激展开激活会演变和变化；c）多次呈现单个刺激不切实际（例如，预期会有重复启动或熟悉效应）。
