返回抑制训练效应的认知神经机制--来自ERP研究的证据

Cognitive neural mechanism of training effect on inhibition of return: Evidence from an ERP study

返回抑制是指个体对呈现在先前(约250 ms前)线索化或注视过位置的刺激反应更慢的现象。尽管有研究表明IOR在长时训练下存在稳定的训练效应,但其相应的认知神经机制,即训练究竟是如何影响IOR的迄今为止尚不明确。本研究采用对信息加工过程高敏感的事件相关电位技术结合长时训练和线索-靶子范式来对该问题进行考察。结果发现:在行为上同前人研究结果相一致,返回抑制效应在长时训练下表现出稳定的下降趋势;更重要的是在脑电上同IOR行为效应量在训练后变小的结果相一致,标识早期知觉加工的成分如N1等在有效和无效线索条件间的差异也表现出在训练后的降低。结果表明早期的知觉加工阶段是训练影响IOR的一个重要阶段,为探明IOR训练效应的认知神经机制提供了来自电生理学的重要证据。

Inhibition of return（IOR） refers to slower responses to targets presented at the previously cued location than to those at uncued locations when the cue-target onset asynchrony is more than about 250 ms. Although much has been debated about whether training could influence IOR, a recent behavioral study（Xu, Ma, Zhang, Zhang, 2015） provided strong evidence for the existing of the IOR training effect. The study observed a reliable and significant decrease in IOR effect under a 8-day sustained training. However, as behavior reflects the combined influence of multiple processing stages, the behavioral measures are unable to determine definitely at which stage the IOR training effect takes place. Thereby, how the training shapes IOR is still an opening question. The current study is aimed to tackle this question by using the Event-Related Potentials that are superior in time-resolution and hence are sensitivity in tracking the distinct information processing stages. 24 paid participants recruited from the campus of Soochow University（6 males, 18 females, mean age of 20.6 ± 2.4, normal or corrected to normal vision） were asked to discriminate target stimuli（＂@＂or ＂＂） presented at either the previously cued（valid condition） or uncued locations（invalid condition） in 9 successive days. During the first and the last training day, the electroencephalogram（EEG） data were acquired while the participants performed the task. The results showed that： 1） Behaviorally, consistent with the results of XU et al.（2015）, the IOR effect（RTvalid- RTinvalid） decreased steadily and significantly as the training days increased（19 and 6 ms for the first and last training day respectively）; 2） Electro-physiologically, compatible with previous ERPs studies of IOR（e.g., Prime Jolicoeur, 2009; Prime Ward, 2006）, the target stimuli occurred at valid locations elicited smaller N1（170~200 ms） as well as the P2 than that at invalid locations; 3） And more interestingly, while N1 cueing effect（invalid- valid） and P2 cueing effect（the Nd250） decreased significantly from the first to the last training day, the P1 cueing effect kept constant across the training days.Taken together, 1） as the visual N1 cueing effect has been well demonstrated to reflect the perceptual processing（e.g., discrimination process）, the current results suggested that perceptual processing is a critical stage during which the training effect of IOR occurs, providing, as far as we known, the first electrophysiological evidence for the cognitive neural mechanism of the training effect of IOR. And 2） the significant regression of the behavioral IOR effect on the N1 and the P2 cueing effect（Nd250） instead of the P1 cueing effect suggested that the N1 and Nd250 may be the more robust and reliable electrophysiological indexes of IOR.