Intermodulation responses show integration of interacting bodies in a new whole

The human visual system is particularly attuned to social cues in the environment (Nakayama, 2010; New, Toby & Cosmides, 2007; Papeo, 2020). Going beyond the study of specialized processes for the perception of social entities such as faces and bodies, recent work has addressed the processing of social interactions (Abassi & Papeo, 2020; Abassi & Papeo, 2022; Bellot et al., 2021; Dima et al. 2022; Isik et al., 2017; Landsiedel et al., 2022; Quadflieg et al., 2015; Quadflieg & Koldewyn, 2017; Tarhan & Konkle, 2020; Walbrin et al., 2018; Walbrin & Koldewyn, 2019; Wurm, Caramazza & Lingnau, 2017; Yang et al., 2015). Some of those studies have shown that, in visual areas specialized to person perception (i.e., face and body processing), perception of social interaction changes the visual representation of individual bodies and body movements (Abassi & Papeo, 2020; Bellot et al., 2021). The representational change registered in visual areas may have a counterpart in behavioral phenomena showing that seemingly interacting bodies (e.g., face-to-face bodies) have faster access to visual attention and awareness, relative to the non-interacting bodies (Papeo et al. 2017; Papeo et al., 2019; Vestner et al., 2019; Yin et al. 2018).

In the present study, we addressed a mechanistic explanation, whereby differences between face-to-face and back-to-back bodies at the level of neural representation and behavioral performance, would reflect the advantage of integrating two independent percepts (two bodies) into a single perceptual unit, or new whole, when perceived as interacting. Research at the neural level has indeed suggested that representing multiple objects as an integrated group increases processing efficiency (e.g., in visual search or visual working memory), by reducing inter-object competition for neural representation and processing resources, and enhancing the neural representation of grouped objects (Kaiser al., 2014; Kaiser et al., 2019).

We used a frequency-tagging electroencephalography (ftEEG) paradigm based on Steady-State Visually Evoked Potentials (SSVEP) to address this hypothesis. SSVEP is a stimulus-locked oscillatory response to a periodic visual stimulation, generated by the neuronal population that responds to that stimulation by oscillating at the same periodicity, the so-called fundamental frequency. Due to non-linear transformations of the signal in neural transmission, the effect at the fundamental frequency can spread over frequencies that are integer multiples of fundamental frequencies, the so-called harmonic frequencies (2f1, 3f2, etc.) (Adrian & Matthews, 1934; Regan, 1966; Regan & Heron, 1969). In humans, responses at fundamental frequencies and harmonics can be measured with the frequency-domain analyses of the recorded EEG signal. In the current study, SSVEP were elicited in female and male human adults, by presenting two items (i.e., two bodies) on the right and left of central fixation, respectively, flickering at two different fundamental frequencies, during EEG recording. Since the two bodies were presented at two different frequencies, we obtained two responses at the corresponding fundamental frequencies F1 and F2, and their harmonics. When two flickering stimuli are simultaneously presented, as in the present paradigm, in addition to the corresponding SSVEPs at F1 and F2, and harmonics, the stimulation can produce a response at the so-called intermodulation (IM) frequencies, which reflects nonlinear interactions between the neural signals associated with the individual items (i.e., sums or differences of individual terms: nF1 ± mF2). The response at IM frequencies is thought to arise from neurons receiving inputs from both stimulations, in which those inputs interact non-linearly, thus indicating the existence and degree of neural integration (Gordon et al., 2019; Norcia et al., 2015; Regan and Regan, 1988). Because IM frequencies do not necessarily overlap with fundamental and harmonic frequencies, they can allow distinguishing neural integration from the response to individual parts (here, single bodies; see Aissani et al., 2011; Alp et al., 2016; Appelbaum et al., 2008; Ratliff & Zemon, 1982; Victor & Conte, 2000; Zhang et al., 2011; Zemon & Ratliff, 1984). We leverage this feature of the SSVEP to measure the distinctive responses to two flickering (facing or non-facing) bodies and their integration.

Effects at IM frequencies triggered by dual stimulation in ftEEG have been extensively studied in vision science, to show for instance feature binding (Aissani et al., 2011) and gestalt perception (i.e., grouping by illusory contours; Alp et al., 2016; grouping by motion synchrony; Alp et al., 2017; figure-ground interaction; Appelbaum et al., 2008). Further research has successfully adapted this method to study integration in higher-level visual processes such as face and biological motion perception (Adibpour et al., 2021; Boremanse et al. 2013; 2014). For example, neural integration could be dissociated from part perception, showing different IM responses for parts (i.e., face halves) that formed a face versus parts that did not form a meaningful configuration (e.g., misaligned face halves) (Boremanse et al. 2013, 2014). Using this approach, the current study sought to capture a dissociation between perception of multiple single bodies and perception of multiple interacting bodies. Dyads of bodies were presented in a spatial relation typical of social interaction (face-to-face), or as unrelated (back-to-back). Individual bodies in a dyad flickered at two different frequencies (the fundamental frequencies F1 and F2). We performed narrowband spectral analysis to distinguish the neural responses to individual bodies (at the fundamental and harmonic frequencies) from the neural effect of integration at IM frequencies nF1±mF2. The specificity of the effects for social stimuli (i.e., bodies) was addressed by also testing objects with a clear anteroposterior morphology (i.e., chairs and machines), which, like bodies, could be presented face-to-face and back-to-back but, unlike bodies, yield no representation of (social) interaction. In addressing the processing of seemingly related (face-to-face) or unrelated (back-to-back) body dyads, the current study investigated how, from visual perception of individual bodies, a transformative visual process proceeds towards the representation of social interaction.

留言 (0)

沒有登入
gif