Brain dynamics of upstream perceptual processes leading to visual object recognition: A high density ERP topographic mapping study

Abstract

Recent studies suggest that visual object recognition is a proactive process through which perceptual evidence accumulates over time before a decision can be made about the object.However, the exact electrophysiological correlates and time-course of this complex process remain unclear. In addition, the potential influence of emotion on this process has not been investigated yet. We recorded high density EEG in healthy adult participants performing a novel perceptual recognition task. For each trial, an initial blurred visual scene was first shown, before the actual content of the stimulus was gradually revealed by progressively adding diagnostic high spatial frequency information. Participants were asked to stop this stimulus sequence as soon as they could correctly perform an animacy judgment task. Behavioral results showed that participants reliably gathered perceptual evidence before recognition. Furthermore, prolonged exploration times were observed for pleasant, relative to either neutral or unpleasant scenes. ERP results showed distinct effects starting 280 ms post-stimulus onset in distant brain regions during stimulus processing, mainly characterized by: (i) a monotonic accumulation of evidence, involving regions of the posterior cingulate cortex/parahippocampal gyrus, and (ii) true categorical recognition effects in medial frontal regions, including the dorsal anterior cingulate cortex. These findings provide evidence for the early involvement, following stimulus onset, of non-overlapping brain networks during proactive processes eventually leading to visual object recognition.

Multiple synergistic effects of emotion and memory on proactive processes leading to scene recognition

“…4A and 4B). Similarly to our previous study 397 (Schettino et al, 2011), we found a reliable topographical change across recognition levels 398 immediately following the exogenous N1 and P2 ERP components (Fig. 4B) 6 .…”

“…Moreover, we replicated our previous results (Schettino et al, 2011) and found that, regardless of 528 the memory status of the scenes (old vs. new), participants dwelt longer on emotional compared 529 to neutral scenes (Fig. 1B).…”

“…We first compared ERP activity for three images 458 before recognition to actual recognition (see Schettino et al, 2011, for a similar approach), 459 separately for new and old scenes. This analysis revealed, for new scenes, a stronger activation 460 for recognition relative to three images before in the dACC (±10x, +15y, +35z) [t(17) = 2.75, p = 461…”

“…Afterwards, the continuous EEG signal was segmented into 258 individual epochs, excluding trials corresponding to errors (based on accuracy on Response2). 259 Similar to our previous study (Schettino et al, 2011), four main epochs were computed around 260 stimulus onset (using an interval of 200 ms prior to and 750 ms after this event), enabling us to 261 look backwards at visual ERPs progressively elicited prior to Response1: (1) segments in which 262…”

“…To examine the ERP data recorded during the progressive unfolding sequence, reference-free 302 topographic analyses were used, similarly to our previous study (Schettino et al, 2011). This 303 method allows to summarize a complex ERP data set into a smaller number of dominant scalp 304 topographies (i.e., global configuration of the electric field across all 128 channels at each time 305 frame) (Lehmann and Skrandies, 1980;Michel and Murray, 2012;Michel et al, 2001Michel et al, , 1999306 Murray et al, 2008;Pourtois et al, 2008).…”