Task relevance effects in electrophysiological brain activity: Early, but not first

Abstract

A current controversy surrounds the question whether high-level features of a stimulus such as its relevance to the current task may affect early attentional processes. According to one view abruptly appearing stimuli gain priority during an initial feedforward processing stage and therefore capture attention even if they are irrelevant to the task. Alternatively, only stimuli that share a relevant property with the target may capture attention of the observer. Here, we used high-density EEG to test whether task relevance may modulate early feedforward brain activity, or whether it only becomes effective once the physical characteristics of the stimulus have been processed. We manipulated task relevance and visual saliency of distracters presented left or right of an upcoming central target. We found that only the relevance of distracters had an effect on manual reaction times to the target. However, the analysis of electrocortical activity revealed three discrete processing stages during which pure effects of distracter saliency (~80-160 ms), followed by an interaction between saliency and relevance (~130-240 ms) and finally pure effects of relevance (~230-370 ms) were observed. Electrical sources of early saliency effects and later relevance effects were localized in the posterior parietal cortex, predominantly over the right hemisphere. These findings support the view that during the initial feedforward stage only physical (bottom-up) factors determine cortical responses to visual stimuli, while top-down effects interfere at later processing stages.

Spatiotemporal brain dynamics underlying attentional bias modifications

“…Ding et al, 2014), a process that is also modulated in a bottom-up fashion by the saliency of the stimuli, even if saliency is irrelevant to the task (Fellrath et al, 2014). The P1 component has also been interpreted as indexing stimulus gating, notably based on evidence for its sensitivity to the attended location (Clark and Hillyard, 1996;Egeth and Yantis, 1997;Eimer and Kiss, 2008;Fellrath et al, 2014;Fichtenholtz et al, 2007;Hillyard and Anllo-Vento, 1998;Luck et al, 2000;Luck et al, 1990;Mangun, 1995;Mangun and Buck, 1998). With regard to this literature, the early ERP modulation observed in our study thus suggests that ABM procedures impact on stimulus gating and on how the stimuli capture attention.…”

“…the bottom-up capture of visual attention by salient stimuli (Clark and Hillyard, 1996;Egeth and Yantis, 1997;Eimer and Kiss, 2008;Fellrath et al, 2014;Fichtenholtz et al, 2007;Hillyard and Anllo-Vento, 1998;Luck et al, 1990;Mangun, 1995;Mangun and Buck, 1998). Three main ERP components manifest over this time period, the C1, P1 and N1.…”

“…; Rauss et al, 2011), this component is also sensitive to stimuli-related internal states, notably including top-down attention (Gilbert and Li, 2013;Kelly et al, 2008) or motivational value (Rossi et al, 2017;Stolarova et al, 2006). The posterior P1 and N1 ERP components peak at 60-190 ms and are sensitive to observer's focus of attention (Fellrath et al, 2014;Mangun, 1995;Mangun and Buck, 1998;Vogel and Luck, 2000; though see e.g. Ding et al, 2014), a process that is also modulated in a bottom-up fashion by the saliency of the stimuli, even if saliency is irrelevant to the task (Fellrath et al, 2014).…”

“…The posterior P1 and N1 ERP components peak at 60-190 ms and are sensitive to observer's focus of attention (Fellrath et al, 2014;Mangun, 1995;Mangun and Buck, 1998;Vogel and Luck, 2000; though see e.g. Ding et al, 2014), a process that is also modulated in a bottom-up fashion by the saliency of the stimuli, even if saliency is irrelevant to the task (Fellrath et al, 2014). The P1 component has also been interpreted as indexing stimulus gating, notably based on evidence for its sensitivity to the attended location (Clark and Hillyard, 1996;Egeth and Yantis, 1997;Eimer and Kiss, 2008;Fellrath et al, 2014;Fichtenholtz et al, 2007;Hillyard and Anllo-Vento, 1998;Luck et al, 2000;Luck et al, 1990;Mangun, 1995;Mangun and Buck, 1998).…”

“…We hypothesized that ABM would affect both early stages of electrocortical processing reflecting gating mechanisms (latencies between 50-100 ms) or bottom-up capture of attention (50 and 200 ms, notably on the P1 parieto-occipital and N2pc ERP component; Fellrath et al, 2014;Hickey et al, 2006) and later, top-down attentional mechanisms (latencies after 150 ms within parietal and frontal sites corresponding to the N2 (Eldar and Bar-Haim, 2010;Osinsky et al, 2014), P2 (Eldar and Bar-Haim, 2010;Suway et al, 2013) and/or P3 components (Eldar and Bar-Haim, 2010)). We further assessed whether modulating attentional biases influenced inhibitory control to the biased stimuli by testing participants in a Go/NoGo task in which they had to withhold responses to the cues biased during ABM training.…”