Does surprise enhancement or repetition suppression explain visual mismatch negativity?

Abstract

A long tradition of electrophysiological studies, using oddball sequences, showed that the neural responses to a given stimulus differ when their presentation occurs frequently (standards) as compared to rare, infrequent presentations (deviants). This difference, originally described in acoustic perception, can also be detected in the visual modality and is termed as visual mismatch negativity (vMMN). Also, a large number of studies detected the reduction of the neuronal response after the repetition of a given stimulus (repetition suppression - RS) and it was suggested that RS is the major mechanism of MMN, an explanation currently also supported by animal studies. However, human studies have proposed that a surprise-related response enhancement for the deviant stimuli might also underlie vMMN. Therefore, the aim of the current study was to disentangle which neural mechanism explains vMMN better: the surprise related response enhancement for the presentation of rare deviants or the RS related to the frequent presentation of the standards. Since the MMN depends strongly on the applied categories, we tested the neural mechanisms of vMMN for different stimulus categories (faces, chairs, real and false characters) using a visual oddball paradigm. We found significant vMMN for every stimulus category. Interestingly, the neural mechanisms behind vMMN were found to be category dependent (assuming no cross-adaptation effects): for faces and chairs it was largely driven by RS, whereas for real and false characters it was mainly due to surprise-related changes.

Immediate stimulus repetition abolishes stimulus expectation and surprise effects in fast periodic visual oddball designs

“…The later part of the response (~330ms in Figure 7B) appears to instead be generated from bilateral posterior sources. Our observations are congruent with expectation effects found over frontal electrodes in EEG (Feuerriegel et al, 2018;Hall et al, in press), and in BOLD signals in frontal areas such as inferior frontal and middle frontal gyri (Grotheer & Kovacs, 2015;Amado et al, 2016) and dorsolateral prefrontal cortex (den Ouden et al, 2009;Rahnev et al, 2011), as well as in ventral temporal regions when presenting face stimuli (e.g., Egner et al, 2010;Grotheer & Kovacs, 2015). Our results also provide further evidence that expectation effects underlying VMRs are not restricted to visual areas, as is commonly assumed in existing studies of visual mismatch responses.…”

“…The finding that expectation effects in visual oddball designs are larger in the presence of stimulus change (i.e., for unrepeated stimuli), is consistent with a substantial body of evidence from non-oddball designs. Previous studies have shown that response differences to expected/surprising stimuli are larger for unrepeated compared to repeated stimuli (Amado et al, 2016;Waconge et al, 2011;Todorovic & de Lange, 2012;Kovacs et al, 2012;Larsson & Smith, 2012;Symonds et al,2016). That stimulus repetition appears to suppress expectation effects suggests that there are at least two interacting mechanisms which underlie visual oddball effects, similar to models proposed to account for auditory mismatch responses (Costa-Faidella, et al, 2011;Mittag et al, 2016).…”

“…Not only are stimulus repetition and expectation effects readily confusable, evidence from other (non-oddball) experiments also indicates that these effects may actually interact. Recent work has found that repetition effects are larger for surprising stimuli, due to large surprise-related signal increases for unrepeated stimuli (Amado & Kovacs, 2016;Larsson & Smith, 2012;Choi et al, 2017;Kovacs et al, 2012Kovacs et al, , 2013reviewed in Kovacs & Vogels, 2014). Similar interaction effects have also been found for ERPs/ERFs, with several studies showing that expectation violation responses can be reduced when stimuli are repeated (Todorovic & de Lange, 2012;Symonds et al, 2017;Wacongne et al, 2011).…”

“…A Deviant ) reveal more negative-going waveforms evoked by deviants at posterior electrodes between 150-300ms (Czigler et al, 2004;Kimura et al, 2009;Stefanics, et al, 2011), an effect known as the visual mismatch negativity (vMMN; for recent reviews see Kimura et al, 2011;Stefanics et al, 2014). The vMMN is considered to be the visual counterpart of the earlier-discovered auditory MMN (Naatanen, et al, 1978), and has been similarly used to investigate a range of phenomena including sensory memory/change detection (Czigler et al, 2002), perceptual discrimination (Tales & Butler, 2006), stimulus repetition effects (Amado & Kovacs, 2016), and perceptual expectations (Stefanics et al, 2014). The magnitude of the visual mismatch response differs between healthy and clinical samples across a wide range of psychiatric and neurological disorders (reviewed in Kremlacek et al, 2016), as has also been found for the auditory MMN (reviewed in Naatanen et al, 2014).…”

“…ERPs to the same stimulus are compared across equiprobable and standard contexts to test for repetition effects, and across equiprobable and deviant contexts to test for expectation/surprise effects (e.g., Amado & Kovacs, 2016; see sequence comparison labels in Figure 1). Results of these studies already suggest that both stimulus repetition effects and expectation effects contribute to the VMR (Amado & Kovacs, 2016;Astikainen et al, 2008;Czigler et al, 2002;Kimura et al, 2009). However, several important confounds persist when using equiprobable control sequences, which can be understood by considering the types of expectations that can be formed during each sequence type.…”