Within-Category Decoding of Information in Different Attentional States in Short-Term Memory

Within-Category Decoding of Information in Different Attentional States in Short-Term Memory
复制标题

DOI:
10.1093/cercor/bhw283
复制
发表时间:
2017-10-01
期刊:
影响因子:
3.7
通讯作者:
Postle, Bradley R.
Postle, Bradley R.
中科院分区:
医学2区
文献类型:
--
作者:
LaRocque, Joshua J.;Riggall, Adam C.;Postle, Bradley R.

文献摘要

被引文献

相似文献

认知神经科学的一个长期假设是工作记忆(WM)是通过持续、升高的神经活动来实现的。最近,WM 理论通过描述具有不同激活水平的 WM 中的不同注意力状态扩展了这一观点。一些研究使用功能磁共振成像 (fMRI) 和脑电图 (EEG) 数据的多变量模式分析 (MVPA) 来研究与这些 WM 状态相对应的神经活动。有趣的是,没有发现 WM 中注意力焦点之外的信息(“无人参与的记忆项目”,UMIs)的主动神经表征的证据,这表明只有有人参与的记忆项目(AMI)才会伴随着活跃的痕迹。然而,这些结果依赖于类别级解码,而类别级解码对单个项目的神经表示缺乏敏感性。因此,我们采用了一项 WM 任务,其中受试者记住两个点阵列的运动方向,并通过回溯指示哪个方向与即将发生的记忆探测(AMI)相关。该设计允许在刺激项级别对延迟周期 fMRI 信号进行 MVPA 解码,从而为 UMI 的神经表征提供更灵敏的测试。虽然 AMI 的证据可靠地很高,但 UMI 的证据却下降到基线,这与不同 WM 注意力状态可能具有性质不同的保留机制的观点一致。
A long-standing assumption of cognitive neuroscience has been that working memory (WM) is accomplished by sustained, elevated neural activity. More recently, theories of WM have expanded this view by describing different attentional states in WM with differing activation levels. Several studies have used multivariate pattern analysis (MVPA) of functional magnetic resonance imaging (fMRI) and electroencephalography (EEG) data to study neural activity corresponding to these WM states. Intriguingly, no evidence was found for active neural representations for information held in WM outside the focus of attention ("unattended memory items,"UMIs),suggesting that only attended memory items (AMIs) are accompanied by an active trace. However, these results depended on category-level decoding, which lacks sensitivity to neural representations of individual items. Therefore, we employed a WM task in which subjects remembered the directions of motion of two dot arrays, with a retrocue indicating which was relevant for an imminent memory probe (the AMI). This design allowed MVPA decoding of delay-period fMRI signal at the stimulus-item level, affording a more sensitive test of the neural representation of UMIs. Whereas evidence for the AMI was reliably high, evidence for the UMI dropped to baseline, consistent with the notion that different WM attentional states may have qualitatively different mechanisms of retention.