The purpose of current study is to reveal spatiotemporal features of oscillatory EEG activities in response to emotional arousal induced by emotional video stimuli, and to find the characteristics of cortical activities showing significant difference according to arousal levels. The EEGs recorded during watching affective video clips were transformed to cortical current density time-series, and then, cluster-based permutation test was applied to determine the spatiotemporal origins of alpha- and beta-band activities showing significant difference between high and low arousal levels. We found stronger desynchronization of alpha-band activities due to higher arousal in visual areas, which may be due to stronger activation for sensory information processing for the highly arousing video stimuli. In precentral and superior parietal regions, the stronger desynchronization in alpha-and low beta-bands was observed for the high arousal stimuli. This is expected to reflect enhanced mirror neuron system activities, which is involved in understanding the intention of other’s action. Similar changes according to arousal level were found also in inter-regional phase synchronization in alpha- and beta-bands.
Citation: Kim H, Seo P, Choi JW, Kim KH (2021) Emotional arousal due to video stimuli reduces local and inter-regional synchronization of oscillatory cortical activities in alpha- and beta-bands. PLoS ONE 16(7): e0255032. https://doi.org/10.1371/journal.pone.0255032
Editor: Vilfredo De Pascalis, La Sapienza University of Rome, ITALY
Received: November 3, 2020; Accepted: July 8, 2021; Published: July 23, 2021
Copyright: © 2021 Kim et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All data and code to generate figures underlying the study are available at Figshare (https://doi.org/10.6084/m9.figshare.14832690).
Funding: This research was supported by the Brain Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science and ICT (2017M3C7A1029485); This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (NRF-2015R1D1A1A01056743).
Competing interests: The authors have declared that no competing interests exist.
Emotional arousal refers to the state of heightened physiological activity, including fear and anger. The state of emotional arousal is highly related to cognitive factors such as attention and memory. Cuthbert et al. reported the correlation between attention and arousal levels while responding to affective pictures . Memory enhancement was observed in response to the stimuli with high arousal [2,3]. Cortical activities due to arousal has been extensively studied using electroencephalogram (EEG) [4–6], and it has been consistently shown that the spectral powers of EEGs in alpha-band decrease in response to arousing stimuli [7–9]. However, contradictory results of alpha-band power increases due to arousal were also reported [10–12].
Most of these studies were performed with affective pictures, whereas video stimuli are more natural and effective to induce emotional arousal. The EEG studies on watching affective video have been rarely reported [13,14], and the characteristics of cortical responses to arousing video stimuli remain unclear. In this study, we tried to reveal spatiotemporal features of oscillatory EEG activities in response to emotional arousal induced by affective video stimuli. Specifically, we investigated the characteristics of cortical activities showing significant difference in EEG subband powers according to arousal levels. This was enabled by combining distributed cortical source model and cluster-based permutation test. In addition, we also explored the inter-regional synchronization/desynchronization of alpha- and beta-band activities. The use of video stimuli was effective since it provides an efficient way to induce emotional arousal, closer to real-world experiences than still image stimuli . The results of this study may eventually contribute to develop a quantitative tool to evaluate the effectiveness of multimedia stimuli, by providing quantitative characterization of cortical responses.
We recorded multichannel EEG signals while the subjects were watching emotional video stimuli. We developed an experimental paradigm that is similar to a real-world live video streaming environment to induce natural emotional responses. Time-frequency analysis and inter-regional functional connectivity analysis were performed, after cortical source reconstruction using distributed source model [16,17]. We used a data-driven approach to find out the features of cortical activities reflecting emotional arousal, based on nonparametric, cluster-based permutation test .
First, we selected 96 affective video clips (48 movie trailers, 48 TV series excerpts, 720p HD) as follows: One-minute length Korean movie trailer clips were selected from YouTube, watched over 1,000 times during 2013 to 2016. TV series excerpts with 2~3 min duration were selected from a Korean video streaming web service (Naver TV, http://tv.naver.com), if watched more than 10,000 times during 2013 to 2016. Then, they were cropped to 1 min duration around the main scene, by visual inspection.
Before the main experiment, twenty-nine subjects (age: 23.2±1.8, 14 males, 15 females) were recruited for the pre-experiment to select the video clips. They assessed valence and arousal levels after watching each video, in 9-point Likert scale using self-assessment manikins (SAM) . From the result of pre-experiment, thirty-two video clips (16 movie trailers, 16 TV series excerpts) were finally selected, by choosing the ones located far away from the x and y axes of the valence-arousal plane as shown in (Fig 1A). That is, the ones with extreme arousal and valence scores.
(A) The distribution of the valence and arousal scores of video clips, averaged over all the participants of the pre-experiment. The finally selected video clips were indicated by shaded ellipses in the valence-arousal plane. (B) Temporal sequence of presenting visual stimuli in a single trial.
Twenty healthy university students were recruited for the main experiment (age: 23.6±1.8, 10 males, 10 females). The subjects were instructed to have enough sleep at the night before the experiment. Overall experiment was divided into four blocks, which consisted of eight trials of ~20 min duration as shown in (Fig 1B). Between each block, the subjects took a rest for ~5 min. Each trial comprised of fixation (30 s), video selection, advertisement (30 s), video presentation after confirmation (1 min), self-assessment (10 s), during which the subjects assessed the levels of valence, arousal, and liking in 9-point Likert scale. The video stimuli were presented by a stimulus presentation software (PRESENTATION; Neurobehavioral systems, Berkeley, CA, USA).
We recorded EEG signals at a sampling rate of 500 samples/s using BrainAmp EEG amplifier (Brain Products, GmbH, Munich, Germany), and an EEG cap with 64 Ag/AgCl electrodes placed according to the international 10–20 system (actiCAP, Brain Products, GmbH, Munich, Germany). The reference and ground electrodes were placed at FCz and AFz, respectively. The contact impedances were kept below 10 kΩ. The signals were recorded with a 60 Hz notch filter, 0.1 Hz highpass filter and 100 Hz lowpass filter. After downsampling (500→200 samples/s), stereotyped artifacts such as eye movement/blink or muscle activities were removed using independent component analysis (ICA, EEGLAB toolbox ). The study has been reviewed and approved by the institutional review board (IRB) of Yonsei University, Wonju campus (1041849-201501-BM-002-01 and 1041849-201804-BM-034-03). A written consent was obtained from all participants.
Data analysis methods
Fig 2 illustrates overall procedure for EEG data analysis. The recorded EEG signals were divided into 90 s epochs (fixation: 30 s, video watching: 60 s). The signals were converted to cortical current source density time-series at 15,002 cortical points. Time-frequency analysis was performed to obtain the powers in alpha-, beta-, theta-, and gamma-bands at each 10 s temporal windows (theta: 4–7 Hz, alpha: 8–12 Hz, beta: 14–20 Hz, gamma: 30–50 Hz). For the functional connectivity analysis, debiased weighted phase lag index (dwPLI) was calculated in each frequency band within each 10 s temporal windows. The spectral powers and dwPLIs were statistically compared between high and low arousal conditions.
The upper and lower parts of the Fig 2 present the procedures for the comparisons of cortical current source strength and inter-regional phase synchronization between high and low arousals.
Cortical current source density estimation.
An open source MATLAB toolbox, Brainstorm , was used to calculate the cortical current source density on the cortical surface. Boundary elements method (BEM) was used for the forward problem. The head model was computed with a symmetric BEM using OpenMEEG implemented in Brainstorm toolbox and the default ICBM152 template with 15,002 points on the cortical surface . The current densities on cortical points were calculated by solving an inverse problem with weighted minimum norm estimation (wMNE) using Brainstorm toolbox .
To reduce the computational load of time-frequency analysis, 15,002 cortical points were downsampled to 500 points (by selecting the process “Less vertices” on the default anatomy in Brainstorm toolbox). And the current densities were projected onto the downsampled brain anatomy (by selecting the process “Project on default anatomy” in Brainstorm toolbox).
The temporal profiles of event-related spectral characteristics were examined using multitaper spectrum estimation (one second long Hanning window, 50% overlap) . The time-frequency distribution of EEG power was calculated as the ratio of the relative power change compared to the baseline value at each frequency in the interval from 10–30 s, in order to reduce intersubject variability and to normalize power changes across different frequencies as follows: (1)
Here Pk(f,t) denotes the power of EEG for the kth trial, estimated at time t and frequency f. R(f) denotes the averaged EEG power within baseline interval at each frequency f. The trial-averaged powers in four frequency bands were obtained for every 10 s interval.
We adopted mass univariate analysis (MUA) for the comparison between two arousal conditions to search the cortical area showing significantly different activation . First, 500 univariate comparisons by t-test were performed for 500 points on cortical surface. Then, the clusters of consecutive points showing high test statistic (t-value) were sought (thresholds: p = 0.0001 for alpha-band, and p = 0.001 for theta-, beta- and gamma-band). If more than five successive cortical points were selected, they were merged to form a cluster. We determined the threshold values for each subband differently so that the cluster size becomes less than 10% of all 500 cortical vertices. If the threshold was set to a specific value without considering the relative size of significant cluster, it occurred that too wide cortical region was determined to be significant, and thus, it became difficult to localize the crucial regions showing significantly different activation between conditions. For example, for alpha-band power, the threshold of t-value was set to smaller (0.0001 = 0.01%) since the clusters tended to be formed too wide.
A cluster was determined to be significantly different between arousal conditions, if its test statistic, the sum of t-values of all cortical points within the cluster, was higher than the threshold. The threshold was determined by surrogate data obtained from 10,000 random permutation of arousal conditions for each subjects , and was set to the highest 0.1% (corresponding to the significance level of 0.001) of the distribution of the test statistic for the surrogate data. FieldTrip toolbox was used for the statistical test .
Functional connectivity analysis.
We conducted functional connectivity analysis on cortical current source time-series, instead of the scalp-level functional connectivity analysis between electrodes which causes poor reliability due to volume conduction/signal leakage . The cortical surface was divided into 62 anatomical region of interests (ROIs) based on the Mindboogle atlas , and then, the current source time-series at the seed cortical point of each ROI was used to perform functional connectivity analysis. The instantaneous phases were calculated by Hilbert transform for the time-series of the four frequency bands (theta, alpha, beta, gamma).
Then, the dwPLI was calculated as a measure of functional connectivity between two cortical points for every 10 s interval for each trial. The dwPLI is a measure of phase synchronization which is weighted for nonzero lag phase activity and debiased for sample size. We adopted the dwPLI since it is known to be much less sensitive to the effect of volume conduction and leakage compared to the phase locking value .
To calculate dwPLI, the weighted phase lag index (wPLI) was calculated first as follows: (2)
Here n denotes the number of time samples within each temporal window. indicates the imaginary part of the cross spectrum between two time-series at time point t. sgn () is signum function. Then, the dwPLI was obtained from the wPLI by debiasing using Vinck’s method . To determine significance at each temporal window, the null distribution of dwPLI was estimated from 10,000 surrogate data, which were generated by randomizing phases of Fourier transform of the time-series . A dwPLI was determined to be significant if it was above the highest 5% of the null distribution.
After subtracting and dividing the baseline values (similarly to the calculation of relative power change in Eq (1)), event-related dwPLI was obtained by averaging over trials. Consequently, positive and negative values of event-related dwPLI indicate the increase and decrease of phase locking during watching video clip relative to baseline, respectively. Paired sample t-test was used to compare the event-related dwPLI averaged over all possible pairs of cortical points (3,844 pairs, from the 62 regions) for each subject, to determine whether the functional connectivity was significantly different between high and low arousal conditions.
Significant differences of self-assessment scores were found between high and low arousal conditions (high: 5.88 ± 1.21, low: 3.33 ± 0.96, t-test, p<0.001), which implies that affective video clips successfully induced emotional arousal as we intended.
Fig 3 shows the distribution of spectral power changes (alpha- and beta-bands) over cortical surface at successive 10 s intervals. The alpha-band power was decreased during watching video stimuli for both high and low arousal conditions, most strongly in visual areas (Fig 3A). For high arousal condition, the alpha-band power decrease was more widespread. The beta-band power was increased in prefrontal and temporal areas, while it was decreased in visual areas during video watching. Prominent decrease of beta-band power was observed over occipital and central regions especially in high arousal condition (Fig 3B).
(A) Alpha-band and (B) beta-band.
The clusters of cortical points showing significant differences in spectral powers between arousal levels were found in alpha- and beta-bands. No significant clusters were found in theta- and gamma- bands. Fig 4 shows the distribution of t-values on cortical surface at each 10 s interval. The t-values above threshold were denoted by blue color (higher spectral power for lower arousal). Significant alpha-band power decreases for higher arousal were found most apparently during 30–40 s, in primary motor cortex, right lateral orbitofrontal, right superior frontal, and left lingual gyrus (Fig 4A). For beta-band power, significant decreases for higher arousal were observed mostly during 20–30 s and 30–40 s intervals, in superior parietal area and precuneus (Fig 4B).
The t-scores within the clusters showing significant difference between high and low arousal conditions were shown by color-codes for (A) alpha-band and (B) beta-band. In total, eight significant clusters were found. More detailed views of the seven significant clusters are depicted in (C) alpha-band and (D) beta-band, respectively.
We found four clusters of cortical points showing significant differences of alpha-band power between high and low arousal conditions (shown in Fig 4C). The largest cluster showing alpha-band power decrease for high arousal condition was found over right superior frontal gyrus in 30–40 s interval (cluster size: 19 cortical points, p<0.001). For the beta-band, we found three clusters showing significant decrease for the higher arousal condition (Fig 4D). The largest cluster was observed at left superior parietal lobule in 30–40 s interval (cluster size: 15 cortical points, p<0.001).
Just as the spectral powers which are expected to represent local synchronization of neural activities, the inter-regional functional connectivity were decreased for high arousal. Fig 5 shows the averaged dwPLIs in alpha- and beta-bands in every 10 s interval. The dwPLIs in both alpha- and beta-bands were significantly decreased for higher arousal at 30–40 second intervals (p<0.05), when the significant differences of alpha- and beta-bands spectral powers were observed. The alpha-band dwPLIs were significantly decreased for higher arousal levels also at other 10 s intervals (p<0.01 for 20–30 s, p<0.05 for 0–10 s, 50–60 s). There was no significant difference between arousal conditions in theta- and gamma-band.
Asterisks denote significant differences between arousal conditions (*: p<0.05; **: p<0.01).
The purpose of this study was to reveal the neural correlates of emotional arousal induced by watching affective video clips. Our results showed that the EEG spectral power and phase synchronization in alpha- and beta-bands were significantly decreased during watching video clips with high arousal, compared to those with low arousal. Our findings may imply that highly arousing video triggers attentional information processing due to incoming visual stimuli, which is reflected in the desynchronization of oscillatory alpha- and beta-band oscillatory activities due to two different mechanisms.
We expect that the stronger desynchronization of alpha-band activities due to higher arousal in visual areas seems to reflect stronger cortical activation for sensory information processing, and is consistent with previous results . The desynchronization of beta-band activities is expected to reflect mirror neuron system activities, which is also stronger for higher arousal.
An important contribution of this study is to characterize the effect of arousal on brain activities in response to emotional video stimuli. Our video stimuli include dynamic visual stimuli and auditory stimuli as well. We observed characteristic brain activities reflecting visual sensory and emotional information processing, which is differential from the responses to still image, although emotional video and still image induces some common cortical activities.
We showed that the alpha- and beta-bands powers and inter-regional phase synchronies are reduced during mid and late temporal periods of watching emotionally arousing video clips over ~1 min period, whereas the still image studies investigate neural activities within several seconds. Despite limitation in spatial resolution and precision, we also tried to reveal the locations on cortical surface showing characteristic changes by distributed source localization. We estimated that the characteristic changes in spectral power and functional connectivity due to arousing videos originate from attentional information processing and mirror neuron system activity.
Alpha-band desynchronization for attentional processing
Alpha-band oscillation has been known to represent attentional inhibitory process which suppresses task-irrelevant information , and has been shown to be decreased in task-relevant regions, whereas increased in task-irrelevant regions [32,33]. This is also in line with our results of the alpha-band desynchronization (Fig 3A), which was focused in visual area. We judge that the stimuli with higher arousal yielded stronger alpha-band desynchronization (Fig 4A), and resulted in enhanced attentional information processing. Previous studies have shown that alpha-band event-related desynchronization in occipital and frontal area is induced in response to emotional picture stimuli . This is in accordance with our results which demonstrated that alpha-band power is reduced in visual area around lingual gyrus and frontal area.
In addition, a number of studies have shown that inter-regional synchronization of alpha-band activities is also decreased due to enhanced attention in cognitive processing [34–36]. Hanslmayr et al. reported that the power and phase synchrony in alpha-band were decreased for enhanced sensory processing due to externally oriented states . Consistently with our results on inter-regional functional connectivity, Cho et al. showed significant decrease of alpha-band phase locking value due to the higher arousal level of rituals .
Alpha- and beta-band desynchronization and mirror neuron system
It is well known that alpha- and low beta-band rhythms are suppressed during action observation and execution [38,39]. Based on this, we estimate that the desynchronization of beta-band activities and that of alpha-band activities in motor-related area may reflect the activation of mirror neuron system (MNS) in response to the video stimuli.
We found that the alpha-band activities over primary motor cortex and beta-band activities in superior parietal region and precuneus were more reduced for higher arousal video in middle-late periods as shown in Fig 4. These results seem to be related to the activation of mirror neuron system, as previous studies suggested that suppression of spectral powers in these frequency bands is a sensitive marker of mirror neuron activity [40–42], which is involved in understanding the intention of other’s action . In addition, it is known that the motion in visual stimuli influences emotional arousal [44,45], and thus, may have induced the activation of cortical areas underpinning action observation. Based on quantitative analysis of motions in video stimuli , we verified that the video clips with higher arousal level contained significantly greater amount of motion contents. Our results on the alpha- and low beta-band power reduction is also in line with a previous study which showed that understanding other’s behaviors or emotions yield power suppression in these frequency bands [47,48].
The alpha- and beta-bands event-related desynchronizations (ERDs) were attributed to the MNS activity, which is known to be related to the observation of movement. Considering that the MNS is also recognized to play a role in understanding other’s intention and higher cognitive function regarding emotional information processing [49–51], our observation of MNS activity change may reflect cortical activity for intention understanding triggered by movement observation, which resulted in emotional arousal. As far as we know, the involvement of MNS in neural response to emotional video and the interpretation of the meaning of this have not been presented in previous studies.
We determined to perform time-series analysis using a finite temporal window because neural response to emotional video stimuli should show dynamic characteristics with wax and wane during overall watching duration. The choice of 10 s window seems to be arbitrary, however, we judged that 10 s is long enough to observe change of cortical activity while short enough to track temporal variation. The point is on the use of time-varying analysis which enabled the tracking of dynamic neural activity change, not the exact size of temporal window. Compared to previous studies which reported the results obtained from an arbitrarily chosen temporal epoch [52,53], we expect that our method provides more valuable information. We have also verified that essentially same results were obtained when we varied the window size to 5 or 15 s.
One of the limitations of the current study is that 64-channel EEGs and standard brain template were used to estimate cortical source activity. The results on the location should be further verified with higher density EEG recording and more rigorous source estimation based on individual MRI, since EEG-based cortical source reconstruction suffers from limited spatial resolution and inaccurate localization. Another limitation is that the scope of this study is only to investigate neural representations for emotional arousal effect. The neural activities reflecting emotional valence or interaction effect between arousal and valence during watching emotional video remain unclear. The questions could be resolved by further studies.
We tried to identify the characteristics of neural activities which show significant differences between high and low arousal during watching emotional videos, including temporal periods and the locations on cortical surface. The alpha-band power reduction due to higher arousal was observed mainly in right superior frontal, orbitofrontal and left lingual gyrus, and this may reflect higher cortical activation for attentional processing. The stronger alpha-band desynchronization in primary motor cortex and the stronger beta-band desynchronization in superior parietal area and precuneus for higher arousal may reflect the increase of mirror neuron system activities. As explained above, all these findings imply that the arousal levels of emotional video induce differentiated neural activities in the mirror neuron system and the cortical areas which are involved in attentional sensory information processing.
- 1. Cuthbert BN, Bradleym MM, Lang PJ. Probing picture perception: Activation and emotion. Psychophysiology. 1996;33(2):103–11. pmid:8851238
- 2. Cahill L, McGaugh JL. Mechanisms of emotional arousal and lasting declarative memory. Trends Neurosci. 1998;21(7):294–9. pmid:9683321
- 3. Hamann S. Cognitive and neural mechanisms of emotional memory. Trends Cogn Sci. 2001;5(9):394–400. pmid:11520704
- 4. Liu Y, Huang H, McGinnis-Deweese M, Keil A, Ding M. Neural Substrate of the Late Positive Potential in Emotional Processing. J Neurosci. 2012;32(42):14563–72. pmid:23077042
- 5. Güntekin B, Başar E. A review of brain oscillations in perception of faces and emotional pictures. Neuropsychologia. 2014;58(1):33–51. pmid:24709570
- 6. Hajcak G, Macnamara A, Olvet DM. Event-related potentials, emotion, and emotion regulation: An integrative review. Dev Neuropsychol. 2010;35(2):129–55. pmid:20390599
- 7. Schubring D, Schupp HT. Affective picture processing: Alpha- and lower beta-band desynchronization reflects emotional arousal. Psychophysiology. 2019;56(8):1–13. pmid:31026079
- 8. de Cesarei A, Codispoti M. Affective modulation of the LPP and α-ERD during picture viewing. Psychophysiology. 2011;48(10):1397–404. pmid:21486291
- 9. Meng X, Liu W, Zhang L, Li X, Yao B, Ding X, et al. EEG oscillation evidences of enhanced susceptibility to emotional stimuli during adolescence. Front Psychol. 2016;7(MAY):1–9. pmid:27242568
- 10. Aftanas LI, Reva N V., Varlamov AA, Pavlov S V., Makhnev VP Analysis of evoked EEG synchronization and desynchronization in conditions of emotional activation in humans: Temporal and topographic characteristics. Neurosci Behav Physiol. 2004;34(8):859–67. pmid:15587817
- 11. Aftanas LI, Varlamov AA, Pavlov S V., Makhnev VP, Reva N V. Time-dependent cortical asymmetries induced by emotional arousal: EEG analysis of event-related synchronization and desynchronization in individually defined frequency bands. Int J Psychophysiol. 2002;44(1):67–82. pmid:11852158
- 12. Uusberg A, Uibo H, Kreegipuu K, Allik J. EEG alpha and cortical inhibition in affective attention. Int J Psychophysiol. 2013;89(1):26–36. pmid:23643563
- 13. Koelstra S, Mühl C, Soleymani M, Lee JS, Yazdani A, Ebrahimi T, et al. DEAP: A database for emotion analysis; Using physiological signals. IEEE Trans Affect Comput. 2012;3(1):18–31.
- 14. Costa T, Rognoni E, Galati D. EEG phase synchronization during emotional response to positive and negative film stimuli. Neurosci Lett. 2006;406(3):159–64. pmid:16942838
- 15. Rottenberg J, Ray RD, Gross JJ. Emotion elicitation using films. Handbook of emotion elicitation and assessment. Handbook of emotion elicitation and assessment Series in affective science. 2007. p. 2–28.
- 16. Kybic J, Clerc M, Abboud T, Faugeras O, Keriven R, Papadopoulo T. A common formalism for the Integral formulations of the forward EEG problem. IEEE Trans Med Imaging. 2005;24(1):12–28. pmid:15638183
- 17. Tadel F, Baillet S, Mosher JC, Pantazis D, Leahy RM. Brainstorm: A user-friendly application for MEG/EEG analysis. Comput Intell Neurosci. 2011;2011. pmid:21584256
- 18. Maris E, Oostenveld R. Nonparametric statistical testing of EEG- and MEG-data. J Neurosci Methods. 2007;164(1):177–90. pmid:17517438
- 19. Morris JD. OBSERVATIONS: SAM: The Self-Assessment Manikin—An Efficient Cross-Cultural Measurement of Emotional Response. J Advert Res. 1995;35(6):63–8.
- 20. Makeig S, Jung TP, Bell AJ, Ghahremani D, Sejnowski TJ. Blind separation of auditory event-related brain responses into independent components. Proc Natl Acad Sci U S A. 1997;94(20):10979–84. pmid:9380745
- 21. Gramfort A, Papadopoulo T, Olivi E, Clerc M. OpenMEEG: opensource software for quasistatic bioelectromagnetics. Biomed Eng Online. 2010;9(1):45. pmid:20819204
- 22. Hämäläinen MS, Ilmoniemi RJ. Interpreting magnetic fields of the brain: minimum norm estimates. Vol. 32, Medical & Biological Engineering & Computing. 1994. p. 35–42. pmid:8182960
- 23. Thomson DJ. Spectrum estimation and harmonic analysis. Proc IEEE. 1982;70(9):1055–96.
- 24. Groppe DM, Urbach TP, Kutas M. Mass univariate analysis of event-related brain potentials/fields I: A critical tutorial review. Psychophysiology. 2011;48(12):1711–25. pmid:21895683
- 25. Helfrich RF, Huang M, Wilson G, Knight RT. Prefrontal cortex modulates posterior alpha oscillations during top-down guided visual perception. Proc Natl Acad Sci. 2017;114(35):9457–62. pmid:28808023
- 26. Oostenveld R, Fries P, Maris E, Schoffelen J-M. FieldTrip: Open Source Software for Advanced Analysis of MEG, EEG, and Invasive Electrophysiological Data. Comput Intell Neurosci. 2011;2011:1–9. pmid:21837235
- 27. Lai M, Demuru M, Hillebrand A, Fraschini M. A comparison between scalp- and source-reconstructed EEG networks. Sci Rep. 2018;8(1):1–8. pmid:29311619
- 28. Klein A, Hirsch J. Mindboggle: a scatterbrained approach to automate brain labeling. Neuroimage. 2005;24(2):261–80. pmid:15627570
- 29. Vindiola MM, Vettel JM, Gordon SM, Franaszczuk PJ, McDowell K. Applying EEG phase synchronization measures to non-linearly coupled neural mass models. J Neurosci Methods. 2014;226:1–14. pmid:24485868
- 30. Vinck M, Oostenveld R, Van Wingerden M, Battaglia F, Pennartz CMA. An improved index of phase-synchronization for electrophysiological data in the presence of volume-conduction, noise and sample-size bias. Neuroimage. 2011;55(4):1548–65. pmid:21276857
- 31. Fell J, Röschke J, Schäffner C. Surrogate data analysis of sleep electroencephalograms reveals evidence for nonlinearity. Biol Cybern. 1996;75(1):85–92. pmid:8765657
- 32. de Pesters A, Coon WG, Brunner P, Gunduz A, Ritaccio AL, Brunet NM, et al. Alpha power indexes task-related networks on large and small scales: A multimodal ECoG study in humans and a non-human primate. Neuroimage. 2016;134:122–31. pmid:27057960
- 33. Neuper C, Pfurtscheller G. Event-related dynamics of cortical rhythms: Frequency-specific features and functional correlates. Int J Psychophysiol. 2001;43(1):41–58. pmid:11742684
- 34. Sauseng P, Klimesch W, Doppelmayr M, Pecherstorfer T, Freunberger R, Hanslmayr S. EEG alpha synchronization and functional coupling during top-down processing in a working memory task. Hum Brain Mapp. 2005;26(2):148–55. pmid:15929084
- 35. Hanslmayr S, Aslan A, Staudigl T, Klimesch W, Herrmann CS, Bäuml KH. Prestimulus oscillations predict visual perception performance between and within subjects. Neuroimage. 2007;37(4):1465–73. pmid:17706433
- 36. Hanslmayr S, Gross J, Klimesch W, Shapiro KL. The role of alpha oscillations in temporal attention. Brain Res Rev. 2011;67(1–2):331–43. pmid:21592583
- 37. Cho PS, Escoffier N, Mao Y, Ching A, Green C, Jong J, et al. Groups and Emotional Arousal Mediate Neural Synchrony and Perceived Ritual Efficacy. Front Psychol. 2018;9:2071. pmid:30416477
- 38. Milston SI, Vanman EJ, Cunnington R. Cognitive empathy and motor activity during observed actions. Neuropsychologia. 2013;51(6):1103–8. pmid:23499724
- 39. Cannon EN, Yoo KH, Vanderwert RE, Ferrari PF, Woodward AL, Fox NA. Action experience, more than observation, influences mu rhythm desynchronization. PLoS One. 2014;9(3):1–8. pmid:24663967
- 40. Puzzo I, Cooper NR, Cantarella S, Russo R. Measuring the effects of manipulating stimulus presentation time on sensorimotor alpha and low beta reactivity during hand movement observation. Neuroimage. 2011;57(4):1358–63. pmid:21664466
- 41. Festante F, Vanderwert RE, Sclafani V, Paukner A, Simpson EA, Suomi SJ, et al. EEG beta desynchronization during hand goal-directed action observation in newborn monkeys and its relation to the emergence of hand motor skills. Dev Cogn Neurosci. 2018;30:142–9. pmid:29524845
- 42. Kilner JM, Marchant JL, Frith CD. Relationship between activity in human primary motor cortex during action observation and the mirror neuron system. PLoS One. 2009;4(3). pmid:19290052
- 43. Iacoboni M, Molnar-Szakacs I, Gallese V, Buccino G, Mazziotta JC. Grasping the intentions of others with one’s own mirror neuron system. PLoS Biol. 2005;3(3):0529–35. pmid:15736981
- 44. Ravaja N. Effects of Image Motion on a Small Screen on Emotion, Attention, and Memory: Moving-Face Versus Static-Face Newscaster. J Broadcast Electron Media. 2004;48(1):108–33.
- 45. Simons RF, Detenber BH, Cuthbert BN, Schwartz DD, Reiss JE. Attention to Television: Alpha Power and Its Relationship to Image Motion and Emotional Content. Media Psychol. 2003;5(3):283–301.
- 46. Hanjalic A, Xu LQ. Affective video content representation and modeling. IEEE Trans Multimed. 2005;7(1):143–54.
- 47. Oberman LM, Hubbard EM, McCleery JP, Altschuler EL, Ramachandran VS, Pineda JA. EEG evidence for mirror neuron dysfunction in autism spectrum disorders. Cogn Brain Res. 2005;24(2):190–8. pmid:15993757
- 48. Woodruff CC, Barbera D, Von Oepen R. Task-related dissociation of EEG β enhancement and suppression. Int J Psychophysiol. 2016;99:18–23. pmid:26593747
- 49. Dapretto M, Davies MS, Pfeifer JH, Scott AA, Sigman M, Bookheimer SY, et al. Understanding emotions in others: Mirror neuron dysfunction in children with autism spectrum disorders. Nat Neurosci. 2006;9(1):28–30. pmid:16327784
- 50. Williams JHG, Waiter GD, Gilchrist A, Perrett DI, Murray AD, Whiten A. Neural mechanisms of imitation and “mirror neuron” functioning in autistic spectrum disorder. Neuropsychologia. 2006;44(4):610–21. pmid:16140346
- 51. Enticott PG, Johnston PJ, Herring SE, Hoy KE, Fitzgerald PB. Mirror neuron activation is associated with facial emotion processing. Neuropsychologia. 2008;46(11):2851–4. pmid:18554670
- 52. Aftanas L, Golosheykin S. Impact of regular meditation practice on EEG activity at rest and during evoked negative emotions. Int J Neurosci. 2005;115(6):893–909. pmid:16019582
- 53. Koelstra S, Yazdani A, Soleymani M, Mühl C, Lee JS, Nijholt A, et al. Single trial classification of EEG and peripheral physiological signals for recognition of emotions induced by music videos. In: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). 2010. p. 89–100.