Department of Experimental Psychology, Ghent University, Belgium.
Department of Experimental Psychology, Ghent University, Belgium; Neuroscience Research Australia, Randwick, Australia; School of Psychology, University of New South Wales, Sydney, Australia.
Neuroimage. 2020 Jul 15;215:116829. doi: 10.1016/j.neuroimage.2020.116829. Epub 2020 Apr 10.
Reward consistently boosts performance in cognitive tasks. Although many different reward manipulations exist, systematic comparisons are lacking. Reward effects on cognitive control are usually studied using monetary incentive delay (MID; cue-related reward information) or stimulus-reward association (SRA; target-related reward information) tasks. While for MID tasks, evidence clearly implicates reward-triggered global increases in proactive control, it is unclear how reward effects arise in SRA tasks, and in how far such mechanisms overlap during task preparation and target processing. Here, we address these questions with simultaneous EEG-fMRI using a Stroop task with four different block types. In addition to MID and SRA blocks, we used an SRA-task modification with reward-irrelevant cues (C-SRA) and regular reward-neutral Stroop-task blocks. Behaviorally, we observed superior performance for all reward conditions compared to Neutral, and more pronounced reward effects in the SRA and C-SRA blocks, compared to MID blocks. The fMRI data showed similar reward effects in value-related areas for events that signaled reward availability (MID cues and (C-)SRA targets), and comparable reward modulations in cognitive-control regions for all targets regardless of block type. This result pattern was echoed by the EEG data, showing clear markers of valuation and cognitive control, which only differed during task preparation, whereas reward-related modulations during target processing were again comparable across block types. Yet, considering only cue-related fMRI data, C-SRA cues triggered preparatory control processes beyond reward-unrelated MID cues, without simultaneous modulations in typical reward areas, implicating enhanced task preparation that is not directly driven by a concurrent neural reward-anticipation response.
奖励能持续提高认知任务的表现。虽然有许多不同的奖励操作,但系统比较却缺乏。在认知控制方面,奖励的作用通常是通过金钱激励延迟(MID;与线索相关的奖励信息)或刺激-奖励关联(SRA;与目标相关的奖励信息)任务来研究的。虽然对于 MID 任务,证据清楚地表明奖励会引发主动控制的全局增加,但不清楚在 SRA 任务中奖励效应是如何产生的,以及在任务准备和目标处理期间这些机制在多大程度上重叠。在这里,我们使用同时进行的 EEG-fMRI 来解决这些问题,使用的是具有四种不同块类型的 Stroop 任务。除了 MID 和 SRA 块之外,我们还使用了带有无关奖励线索的 SRA 任务修改(C-SRA)和常规的奖励中性 Stroop 任务块。行为上,与中性相比,我们观察到所有奖励条件下的表现都更好,与 MID 块相比,SRA 和 C-SRA 块的奖励效应更为明显。fMRI 数据显示,在表示奖励可用性的事件中(MID 线索和(C-)SRA 目标),与价值相关的区域中存在相似的奖励效应,并且无论块类型如何,所有目标的认知控制区域都存在相似的奖励调节。EEG 数据也反映了这种结果模式,显示了明确的估值和认知控制标记,这些标记仅在任务准备期间有所不同,而在目标处理过程中,奖励相关的调节在不同的块类型之间再次具有可比性。然而,仅考虑与线索相关的 fMRI 数据,C-SRA 线索引发了与奖励无关的 MID 线索之外的预备控制过程,而没有在典型的奖励区域中同时进行调节,这意味着增强了任务准备,但不是由同时的神经奖励预期反应直接驱动的。