香京julia种子在线播放

The power of music over human emotions is intriguing and there is an ongoing debate regarding not the mechanisms of how music can provoke pleasure but rather why music can be a rewarding experience (Goupil and Aucouturier, 2019). Music causes both marked changes in an individual’s emotional state to the point of musical thrill (Goldstein, 1980) and collective emotional contagion in a social context (Egermann et al., 2011). At the individual level, intense musical emotion involving areas of the limbic system can provoke a pleasurable rewarding experience: the musical chill (Blood and Zatorre, 2001; Salimpoor et al., 2009, 2011). In their paradigm, the participants provided highly pleasurable musical extracts and continuously reported their felt emotional pleasure while listening to music (neutral, low pleasure, high pleasure, and chills). The phenomenon of pleasure associated with chills is composed of two phases, an anticipation phase before the chill when the pleasure is growing, with a specific dopamine release in the dorsal striatum (caudate), and a peak pleasure phase with a dopaminergic release in the ventral striatum (nucleus accumbens) (Salimpoor et al., 2011). In addition, a pharmacological study by Ferreri et al. (2019) demonstrated that dopaminergic releases were not only the consequence but actually one of the causes of the felt emotional pleasure. These cerebral investigations remain limited to laboratory experiments with heavy neuroimaging techniques (fMRI, PET scan) whereas social neurosciences, and consequently the study of collective emotions, is moving toward natural/ecological paradigms (Acquadro et al., 2016; Volpe et al., 2016; Dikker et al., 2017; Bevilacqua et al., 2018; Soto et al., 2018; Swarbrick et al., 2018; Matusz et al., 2019). The use of mobile wireless EEG could provide the possibility of studying cerebral activity during peak emotion of musical chills in ecological/natural conditions, especially because it can be used in hyperscanning paradigms with several participants simultaneously (Lindenberger et al., 2009; Müller et al., 2013, 2018; Sänger et al., 2013; Acquadro et al., 2016).

Electroencephalography (EEG) is a direct measure of electrical activity of the brain that confers a high temporal resolution in the millisecond range. With high-density EEG (HD-EEG), a large area of the scalp is covered by electrodes, which makes it possible to investigate the cortical sources of the surface activity with an acceptable spatial resolution (estimated in the range of a centimeter). Although musical chills have not yet been studied by EEG, HD-EEG enables the source reconstruction to identify the cortical origins of surface activities (Michel and Brunet, 2019). The activities of these sources can therefore be compared with activities already known in structures implicated in musical reward.

Earlier data from classic EEGs suggest that low frequencies such as theta oscillations are involved in reward processing (Aftanas and Golocheikine, 2001; Wingerden et al., 2010; Gruber et al., 2013). Theta activity is reported to be linked to memory processing in reward contexts (Gruber et al., 2013; Pu and Yu, 2019). A stronger activity in the theta frequency range was found in fronto-medial sites after rewarding feedback onset (Pornpattananangkul and Nusslock, 2016) and at occipito-parietal and fronto-central electrodes when comparing anticipation of high versus low reward (Steiger and Bunzeck, 2017). In the musical context, theta band changes in the fronto-central area during music appreciation have been reported and also correspond to high emotional arousal (Lin et al., 2010). Compared with neutral musical excerpts, highly pleasant music provokes a higher oscillatory activity in the theta band over fronto-central regions than neutral musical excerpts (Nemati et al., 2019). The theta power increases in the medial frontal area for positive musical stimuli (Omigie et al., 2015) and for pleasant music (Sammler et al., 2007). The more the loops of the anterior cingulate cortex (ACC) and the medial frontal cortex are activated while listening to music, the more the fronto-medial theta power increases (Jäncke et al., 2015). The specific theta/alpha activities relative to the valence of the music (both aversive/negative and attractive/positive) might be mediated by the amygdala, which is a direct modulator of the auditory and orbitofrontal cortex for processing musical emotion (Omigie et al., 2015). It has also been suggested that theta oscillations play a role in the synchronization of temporal and prefrontal structures during musical emotional processing (Lewis, 2005). Furthermore, the overall frontal alpha EEG asymmetry (power difference between right and left hemisphere), which is reported to be a good indicator of the emotional state while listening to music (Schmidt and Trainor, 2001; Davidson, 2004; Harmon-Jones et al., 2010; Ramirez and Vamvakousis, 2012; Vecchiato et al., 2012; Arjmand et al., 2017; Mennella et al., 2017; Ramirez et al., 2018), should be an interesting candidate for the study of cortical patterns involved during musical chills.

The aim of the present study was to use HD-EEG to identify specific cortical patterns that underlie musical chills. Our main hypothesis was that an increased felt pleasure on a scale from neutral (lowest intensity) to chills (maximum intensity) should produce an increase of the theta activity over the fronto-central regions, and a specific pattern of theta activity over the centro-parietal and temporal regions.

The aim of the present study was to investigate the neural processing that occurs during musical chills using HD-EEG. The EEG results confirmed a specific activity over the prefrontal regions in the theta frequency band, but we observed no specific alpha asymmetry patterns relating to the approach/withdrawal system. Source localization allowed us to identify greater activities in the insula, OFC, and SMA, which confirmed the specificity of surface activity for chills. These three activated areas are in line with previous findings of neuroimaging studies that also identified the bilateral insula, OFC, ventromedial prefrontal cortex, SMA, and ventral/dorsal striatum involved in musical chills (Blood et al., 1999; Blood and Zatorre, 2001; and Brown et al., 2004).

The specific theta activity has been previously identified in musical emotional processing (Ramos and Corsi-Cabrera, 1989). The results from the prefrontal areas confirm previous findings that suggested a higher theta band activity over the fronto-midline regions and fronto-medial region for pleasant versus neutral musical excerpts (Sammler et al., 2007; Jäncke et al., 2015; Nemati et al., 2019), for positive musical stimuli (Omigie et al., 2015; Rogenmoser et al., 2016). Theta activity might be linked with reward anticipation, memory, and attention (Burgess and Gruzelier, 1997; Krause et al., 1999; Gruber et al., 2013; Pu and Yu, 2019), which is a key component of emotional responses to music (Brattico and Pearce, 2013). More precisely, theta activity is associated with successful memory performance in context of high reward and is reported to be linked with dopaminergic activity (Gruber et al., 2013). The specific frontal theta activity could be the surface activity reflection of the reward system activation via a network including the amygdala, insula, and OFC, which has previously been found to correspond to the occurrence of chills (Blood and Zatorre, 2001). The representation of source activity also suggests that the ventral tegmental area is involved during the chills, despite it not being one of the structures we targeted in our analysis. The OFC is involved in reward processing (Berridge, 2003) and is strongly connected with reward structures such as the amygdala, parahippocampal cortex, and medial prefrontal cortex (Bechara et al., 1994). It is implicated in the system underlying autonomic responses to music (Koelsch, 2014), and activation of the right OFC has been correlated with pleasantness ratings (Blood et al., 1999) and intensity of chills (Blood and Zatorre, 2001). It has been established that the intensity of chills is correlated with the intensity of striatum and OFC activity (Blood and Zatorre, 2001; Salimpoor et al., 2011) and that theta activity could be a cortical activity related to reward structure activation. Considering these facts in addition to the gradual increase of theta activity in the prefrontal ROI when the emotional ratings increased, we can reasonably hypothesize that the increased power of the theta activity is linked to the intensity of the felt pleasure. Further theta activity over fronto-central sites might be dependent on arousal level (Sammler et al., 2007). As suggested by works from Jäncke et al. (2015), the increased activity in the right versus left OFC that was observed with source localization corresponded to the significant right prefrontal increased theta activity. It is also one more argument in favor of the implication of the OFC in the production of theta prefrontal activity. However, the same overall trends found in both the left and right prefrontal areas suggest that musical pleasure is reflected on both left and right frontal sites, independent of positive or negative emotional valence.

The SMA is known to be associated with the reward system (Knutson et al., 2001), and its involvement has previously been identified during musical chills using PET; the rCBF increased in the SMA during chills (Blood and Zatorre, 2001). Furthermore, a correlation analysis demonstrated that the rCBF in the SMA (as well as the insula and the OFC) was positively correlated with increase in several psychophysiological parameters (HR, respiration rate, and electromyography). Decreased theta oscillatory activity in the central region may be related to the specific SMA activity during chills. Both source localization and surface activity showed a large difference for chills compared with the two other conditions, and there were no differences between low versus high pleasure. At this point, we are not able to explain the stronger effect of theta decrease in the right compared with the left central ROI, even though—as for the prefrontal areas—the same trends were found in both left and right ROIs.

The results of our study also highlighted a gradual decrease of theta power for the right temporal ROI and an increased activity in both the left and the right superior temporal gyrus. Because the superior temporal gyrus is known to be activated in the processing of musical liking/disliking (Brown et al., 2004), and familiarity influences neural responses in the auditory cortex (Pereira et al., 2011), we hypothesized that this temporal activity is related to the auditory cortex. The auditory cortex localized in the superior temporal gyrus is connected to the OFC, and these two structures exchange information during music processing (Brown et al., 2004). The activity, which was identified with source localization from the right superior temporal gyrus by Joucla et al. (2018), is likely to be a signature of music liking, and these findings seem to be consistent with ours. Furthermore, a study predicted that people who frequently experience musical chills have a higher structural connectivity (larger volume of white matter connectivity) between the prefrontal medial cortex, insula, and posterior superior temporal gyri (Sachs et al., 2016). We suggest that the theta activity in the right temporal ROI could be related to music liking processing. However, we observed increased activation of both the bilateral insula and the left and right temporal gyri with higher emotional ratings. It is difficult to draw a clear conclusion about the right temporal surface activity because the insula and the auditory cortex are adjacent structures that could both have been involved.

We found no specific effect both for oscillatory rhythms in the alpha band or for alpha frontal and prefrontal asymmetry. We hypothesized that the climax of an emotional increase—related to the activation of the approach dimension relative to positive/rewarding stimulation (Omigie et al., 2015)—from neutral to chills should gradually activate the approach dimension, resulting in increased alpha asymmetry. The extracts for each participant were not restricted to only positive or negative valence, and this should have affected the asymmetry results. The major or minor mode of the music can transmit happy or sad emotions, which, respectively, increase or decrease the alpha power in the left frontal area (Schmidt and Trainor, 2001; Tsang et al., 2001). The experience of chills is quite difficult to reproduce on demand so we did not restrict the valence of the stimuli because this was neither an aim nor would have been a gain for these investigations. We considered that such a restriction was not in line with our perspectives and would have reflected the cerebral activity relative to only positively or negatively valenced pleasurable musical chills.

The arousal calculation showed an increased β/α ratio on frontal regions. There was no effect on the alpha band over the frontal region; however, the same non-significant trends can be identified for most of the participants, who exhibited decreased alpha power on the frontal ROI during chills, which explains our results. Several structures were identified using fMRI, such as the amygdala, prefrontal cortex, or even auditory cortex, which were linked with arousal changes while listening to music (Daly et al., 2019). Further connectivity analysis using fMRI has already demonstrated a close relationship between the nucleus accumbens, ventral tegmental area, and hypothalamus in investigations of affective responses to music (Menon and Levitin, 2005). Connectivity analysis also highlighted a close relationship between several structures of the affective and autonomic systems. The hypothalamus and particularly the insula, which showed connectivity with the nucleus accumbens during music listening (Menon and Levitin, 2005), corresponds to the sources localized in our study.

There are some limitations to our experiment. Because of experimental constraints, the chills might have been “auto-induced” because participants expected the specific chill-inducing moment that they had previously indicated. This could have reinforced the anticipation phase highlighted by Salimpoor et al. (2011). Thus, the investigation of the chills induced by music in a natural ecological setting could provide somewhat different EEG results. Furthermore, several participants reported that the “aseptic” environment of our experimental room could influence the emotions, and thus the experience of chills. For this reason, some participants did not report enough chills and were therefore excluded from EEG analysis. More occurrences of chills per subject, as well as a larger sample, would have reinforced the overall EEG results and the robustness of effects. The influence of the headphones on the local electric field has not been evaluated and the motor activity related to continuous reports might have slightly influenced the oscillatory activity.

In conclusion, our results suggest that HD-EEG could provide relevant information about musical pleasure and pleasurable musical chills mainly in the right prefrontal, temporal, and central areas in the theta band. This work is a first step for the study of musical chills using EEG in more ecological paradigms. In future experiments, with the development of wireless mobile EEG systems and coupled with physiological measures, EEG recordings characterizing musical pleasure and musical chills in both social contexts and live conditions would provide a relevant parameter to study emotional synchronization of groups (Chabin et al., 2020).

The datasets presented in this study can be found in online repositories at the following address: https://figshare.com/articles/Cerebral_activity_during_peak_emotion_in_response_to_music_revealed_by_High-Density_EEG_/11687868.

The study was approved by an independent ethics committee (CPP Ouest V – Rennes; no. 2018-A01653-52) and follows recommendations from the French Jardé’s Law (Article R1121-1 1 of the French Law Code of Public Health amended by decree 127 no. 2017-884 of May 2017) on non-invasive protocols involving healthy humans. All participants received the information in full both orally and on paper, and signed a written informed consent form before inclusion in the study.

TChab, AC, DG, and LP: conceptualization. TChab, CJ, LM, and LP: investigations. TChab and TChan: data curation. TChab and DG: formal analysis. TChab, DG, and LP: writing–original draft. TChab, AC, DG, EH, TM, and LP: writing–review and editing. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.