c o r t e x 6 0 ( 2 0 1 4 ) 8 2 e9 3

Available online at www.sciencedirect.com

ScienceDirect
Journal homepage: www.elsevier.com/locate/cortex

Special issue: Research report

How functional coupling between the auditory

cortex and the amygdala induces musical emotion:
A single case study

Catherine Li
egeois-Chauvel a,b,*, Christian B
enar a,b, Julien Krieg a,b,

Charles Delb
e c, Patrick Chauvel a,b,d, Bernard Giusiano a,b,d and
Emmanuel Bigand c
a
INS INSERM, UMR U, 1106 Marseilles, France
b
Aix-Marseille Universit
e, 13005 Marseilles, France
c
LEAD UMR 5022 CNRS, Universit
e de Bourgogne, 21065 Dijon, France
d
H^opitaux de Marseille, H^opital de la Timone, 13005 Marseille, France

article info abstract

Article history: Music is a sound structure of remarkable acoustical and temporal complexity. Although it
Received 30 May 2013 cannot denote specific meaning, it is one of the most potent and universal stimuli for
Reviewed 8 December 2013 inducing mood. How the auditory and limbic systems interact, and whether this interac-
Revised 3 June 2014 tion is lateralized when feeling emotions related to music, remains unclear. We studied the
Accepted 4 June 2014 functional correlation between the auditory cortex (AC) and amygdala (AMY) through
Published online 16 June 2014 intracerebral recordings from both hemispheres in a single patient while she listened
attentively to musical excerpts, which we compared to passive listening of a sequence of
Keywords: pure tones. While the left primary and secondary auditory cortices (PAC and SAC) showed
Human larger increases in gamma-band responses than the right side, only the right side showed
Auditory cortex emotion-modulated gamma oscillatory activity. An intra- and inter-hemisphere correla-
Amygdala tion was observed between the auditory areas and AMY during the delivery of a sequence
Emotion of pure tones. In contrast, a strikingly right-lateralized functional network between the AC
Music and the AMY was observed to be related to the musical excerpts the patient experienced as
happy, sad and peaceful. Interestingly, excerpts experienced as angry, which the patient
disliked, were associated with widespread de-correlation between all the structures. These
results suggest that the right auditoryelimbic interactions result from the formation of
oscillatory networks that bind the activities of the network nodes into coherence patterns,
resulting in the emergence of a feeling.
© 2014 Elsevier Ltd. All rights reserved.

* Corresponding author. INSERM, UMR U 1106, Faculte
de Me

decine, 27, Bd Jean Moulin, 13005 Marseille, France.

geois-Chauvel).
E-mail address: catherine.liegeois-chauvel@univ-amu.fr (C. Lie
http://dx.doi.org/10.1016/j.cortex.2014.06.002
0010-9452/© 2014 Elsevier Ltd. All rights reserved.
 c o r t e x 6 0 ( 2 0 1 4 ) 8 2 e9 3 83

reward/motivation situations. These regions include limbic

1. Introduction [amygdala (AMY)] and paralimbic structures (orbito frontal
cortex, parahippocampal gyrus), as well as the medio-
Music is a sound structure of remarkable acoustical and
ventral prefrontal cortex, anterior cingulate cortex and the
temporal complexity and although it cannot denote specific
ventral striatum (Blood & Zatorre, 2001; Levitin & Menon,
meaning, it is one of the most potent and universal stimuli
2003; Menon & Levitin, 2005; Gosselin et al., 2006; Pereira
for inducing mood, evoking comparable emotional re-
et al., 2011; Schmithorst & Holland, 2003). Strong AMY acti-
sponses across different musical categories and cultures
vation is reported for negative emotions (Koelsch, Frotz, von
(Peretz & Hebert, 2000; Fritz et al., 2009). The induced affec-
Cramon, Muller & Friederici, 2006; Mitterschiffthaler, Fu,
tive states can be strong and long-lasting (Krumhansl, 1997)
Dalton, Andrew, & Williams, 2007) and for both positive
and very short stimuli may be sufficient to trigger emotional
and negative stimuli (Eldar, Ganor, Admon, Bleich, &
responses (up to 250 msec, Bigand, Filipic, & Lalitte, 2005).
Hendler, 2007). Yet, the modulation of AMY activity is
Understanding the underlying mechanism of how emotion
more likely to be observed during a subject's emotional ex-
is induced by listening to music remains a matter of study
periences, or when emotional music is combined with either
tracking down the underlying mechanisms. Juslin and col-
movies or affective pictures, suggesting that the involve-
laborators (Juslin, 2013; Juslin & Va € stfja
€ ll, 2008) have pro-
ment of this region depends on the context of the subjective
posed a comprehensive framework taking into account
emotional experience associated with the stimuli (Blood &
several physiological mechanisms underlying emotions
Zatorre, 1999; Eldar et al., 2007; Baumgartner, Esslen, &
induced by music, the BRECVEMA model (Brainstem re- € ncke, 2006).
Ja
flexes-Rythmic entrainment, Evaluative conditioning-
Up to now, few studies have investigated the neural
emotional Contagion, Visual imagery-Episodic memory-
mechanisms underlying how acoustical features of stimuli
Musical expectancy, Aesthetic judgment). In addition, the
modulate emotional feeling by looking at the functional
phenomenon of entrainment should be mentioned as an
coupling between the AC and AMY. Recently, Kumar,
important candidate for emotion induction through music,
Stephan, Warren, Friston & Griffiths (2012) showed that
with recent empirical evidence showing a strong link be-
aversive sounds are first processed in the AC, which can
tween the feeling of entrainment and emotion (Labbe
&
foresee the assignment of valence in the AMY. The AMY, in
Grandjean, in press). From another angle, it is largely
turn, can modulate the AC in accordance with the valence of
acknowledged that the acoustic feature content in musical
the sounds. The AMY has extensive connections with many
excerpts covaries with its emotional valence. Converging
cortical and subcortical areas, including back projections to
evidence shows that musical features such as the tempo
the AC, that may modulate sensory processing in these areas
(speed of the beat in music) and the mode (major or minor)
based on emotional signals (Amaral & Price, 1985; Price, 2003;
are relevant in evoking a sense of happiness or sadness
Hoistad & Barbas, 2008). This is in line with studies showing
(Dalla Bella et al., 2001; Khalfa, Scho € n, Anton & Lie
geois-
the increase of AC activation in response to pleasant and/or
Chauvel, 2005) and are associated with specific brain signa-
unpleasant sounds compared to neutral sounds, which
tures (Trochidis & Bigand, 2013). Numerous studies have
would suggest this modulation could be primed by the AMY
contributed to the identification of brain areas in the right
through re-entrant projections (Grandjean et al., 2005; Plichta
and left hemispheres that seem likely to support some stage
et al., 2011).
of the perceptive or emotional processing of music (for re-
Emotion induced by music is a dynamic process; therefore,
views see, Peretz & Zatorre, 2005; Stewart, Von Kriegsten,
its evolution over time in relation to listening is more precisely
Warren & Griffiths, 2006).
tracked by electrophysiological recordings than functional
With respect to auditory processing, it has been argued
imagery. Simultaneous and bilateral depth electrode re-
that auditory cortices in the two hemispheres are relatively
cordings from both AC and AMY in a single epileptic patient
specialized: the right auditory cortex (AC) underlies fine-
during presurgical evaluation, allowed us to study the neural
grained pitch (frequency) processing, whereas the left AC
activity underlying this process with great spatio-temporal
processes rapidly changing broadband stimuli (Lie
geois-
precision. We investigated the functional coupling between
Chauvel, de Graaf, Laguitton, & Chauvel, 1999; Lie
geois-
these structures by applying nonlinear correlations of intra-
Chauvel, Giraud, Badier, Marquis, & Chauvel, 2001; Warrier &
cerebral EEG while the patient listened either to musical ex-
Zatorre, 2004; Zatorre & Belin, 2001; Zatorre, Belin, &
cerpts conveying different emotions (such as happiness,
Penhune, 2002). In addition, brain imaging, electrophysiolog-
sadness, peacefulness & anger) or to a sequence of pure tones
ical and neuropsychological studies have established that
(neutral stimuli). Although emotion evoked by music may be
secondary AC (particularly in the right hemisphere) is asso-
analyzed in numerous ways (see Zentner, Grandjean, &
ciated with the processing of contour properties of unfamiliar
Scherer, 2008 for a full account), in the present study, we
melodies (Behne, Scheich, & Brechmann, 2005; Brechmann &
used a more convenient and traditional way to evaluate the
Scheich, 2005; Brattico, Tervaniemi, Na €a€ ta
€ nen, & Peretz,
feelings experienced by the patient, which was more appro-
2006). In contrast, primary AC (PAC) only processes features
priate for the clinical setting.
of isolated sounds (Samson & Zatorre, 1988; Johnsrude,
The rationale behind this approach was to examine
Penhune, & Zatorre, 2000; Patterson, Uppenkamp, Johnsrude
whether the ACeAMY coupling: (i) is influenced by the
&Griffith, 2002).
emotional valence judgments or by the complexity of the
Music-evoked emotional activations involve brain regions
stimuli (music vs pure tones), and (ii) is modulated as a
known to also be activated by emotional pictures and
function of hemisphere.
84 c o r t e x 6 0 ( 2 0 1 4 ) 8 2 e9 3
2. Experimental procedures
2.1. Subject
CV, a 45-year-old French-speaking woman with a draughts-
manship's degree (10 years of education), participated in this
study. She suffered from drug-resistant partial epilepsy and
was implanted for pre-surgical investigation with chronic
depth electrodes (Stereotactic EEG, SEEG) in the right and left
AC [i.e., Heschl's gyrus (primary and secondary AC (PAC and
SAC) and BA 22)], as well as in the right and left AMYe (Fig. 1,
see Supplemental Material for details about the anatomical
reconstruction of electrodes). These structures were not
involved in the epileptogenic network.
The choice of the anatomical location of electrodes was
based on clinical and video-EEG recordings, as well as MRI,
and was made independently of the present study. This pa-
tient provided informed consent to the protocol, approved by
the Institutional Review Board of the French Institute of
Health. A standard Oldfield test indicated that she was
ambidextrous. She displayed a left hemisphere specialization
for language and normal memory scores (Wechsler Memory
scale III, QM ¼ 100). Recordings of brainstem evoked potentials
and pure-tone audiograms carried out before SEEG indicated
intact cochlear and brainstem auditory functions.
2.2. Stimuli
2.2.1. -Pure tones
Auditory evoked potential recordings are part of the func-
tional mapping procedure that aims to characterize different

geois-Chauvel, Musolino, & Chauvel, 1991;
auditory areas (Lie

geois-Chauvel, Musolino, Badier, Marquis, & Chauvel,
Lie
1994); they were recorded before the musical procedure in
this study. 420 tone bursts of four tonal frequencies (500 Hz,
1 Khz, 2 Khz, 3 Khz) were equally and randomly delivered at a
rate between 800 msec and 1.2 sec. Each tone burst had a
0.3 msec rise and decay time and a total duration of 30 msec.
2.2.2. Musical excerpts
Forty musical excerpts of 20 sec duration were selected by
music theorists and psychologists in order to elicit a homo-
geneous emotional experience. The excerpts (some examples
are available online: http://leadserv.u-bourgogne.fr/wave/)
pertain to the four main musical periods of Western classical
music (i.e., baroque, classic, romantic and modern) and
involve varied instrumental groups (solo, chamber music,
orchestra). Most of them were shown to involve clear emotion,
belonging to the four main categories: happy, sad, angry or
peaceful (Bigand et al. 2005). Furthermore, their emotional
qualities have been confirmed by behavioral measures in
numerous studies, such as Khalfa et al. (2008a, b) or Filipic,
Tillman & Bigand (2010).
2.3. Experimental procedure
During the recording session, the patient, comfortably seated
in an armchair, was instructed to listen and to concentrate on
her own feelings without reporting her emotional response. A
follow-up behavioral study was conducted later in order to
assess the patient's own emotional classification between the
given four emotions and her feelings (Liking or disliking) for
each excerpt using a scale from 1 to 10. We have chosen this
strategy because explicit instructions to monitor and report
feelings could interfere with the emotional reactions one is
attempting to assess.
The stimuli were presented in pseudo-randomized order
using E-prime 1.1 (Psychology Software Tools Inc. Pittsburgh,
PA, USA), through two free field Yamaha NS10M loudspeakers.
The melodies were delivered during 20 sec with an ISI of 12 s.
Data acquisition started 3 min before the presentation of the
first melody in order to have an EEG resting-reference period.
2.4. Electrophysiological recordings
The recordings of intracerebral EEG were monopolar, with
each contact of a given depth electrode referenced to an extra-
dural lead using acquisition software and a 256-channel
BrainAmps EEG amplification system (Brain Products GmbH,
Munich, Germany). During the acquisition the EEG signal was
amplified with a digital band-pass filter (.5e200 Hz) and digi-
tized at a rate of 1 KHz per channel (resolution: 1 msec per
sample).
2.5. Data analysis
2.5.1. Time frequency analysis
In order to analyze the reactivity of cortical structures to pure
tones and to music in relation to emotion, we performed time-
frequency analysis on data in a bipolar montage using Matlab
software (MathWorks, Natick, MA). A wavelet transform
(Gabor, oscillation parameter x ¼ 7) was performed on the
whole temporal window (30 sec), from 10 sec before the onset
of the melody through the 20 sec duration of the excerpt, for
each trial. Then we averaged the time-frequency energy maps
across trials. We performed statistical quantification of the
changes between stimulus and baseline periods, or between
stimulus periods across emotions. We used a non-parametric
ManneWhitney Wilcoxon test, with a threshold of signifi-
cance set to .05. In the case of pure tones, we used the same
pool of data (20 trials of 10 sec of silence) for the baseline
(resting state).
To compare the four emotional categories, we used two
models: ‘power ~ cortical area * emotion * frequency band’,
and for each frequency band ‘power ~ cortical area * emotion’
(cortical areas recorded from 14 contacts). The power data
have been normalized by a logarithmic and arc-tangent
transformation. Each group of post hoc contrasts has been
corrected by the Holm-Bonferroni method. We kept only the
most significant differences to account for the overall
correction of these multiple comparisons.
2.5.2. Estimation of nonlinear correlations between SEEG
signals
The functional couplings between SEEG signals recorded from
distinct cortical structures, auditory areas and AMY were
estimated using nonlinear regression analysis (Lopes da Silva,
Pijn, & Boeijinga, 1989; Wendling, Bartolomei, Bellanger, &
 c o r t e x 6 0 ( 2 0 1 4 ) 8 2 e9 3 85

Fig. 1 e Anatomical reconstruction of the intracerebral electrodes localized in the right amygdale (A) and AC (H and T) and
homologous regions in the left hemisphere (A′, H′ and T′): The first three leads of A and A′ recorded activity from AMY
electrodes; H and H′ recorded activity from the primary AC (PAC) and the secondary AC (SAC) according to the following leads: H
1,2,3 and H’ 5,6,7 recorded activity from the right and left PAC, respectively; the leads H 4 to H9 and H′ 8 to H′ 13 were in right and
left SAC; and T and T’ recorded activity from the associative auditory areas, BA 22: T1 to T4 and T′10 to T′ 14.

Chauvel, 2001; Wendling, Bartolomei, Bellanger, Bourien, &
Chauvel, 2003)
Nonlinear regression analysis is aimed at quantifying the
relationship (or association) between two SEEG signals X(t) and
Y(t) recorded from two neuronal populations Px and Py, without
making any assumptions regarding the nature (linear or
nonlinear) of this relationship (for details, see Wendling et al.,
2001). For the sake of simplicity, the nonlinear correlation co-
efficient h2* xy will be simply denoted as h2. Here, we consider
the maximum value between h2 (X- > Y) and h2 (Y- > X).
In order to limit the influence of the common reference,
bipolar signals (obtained from the subtraction of monopolar
signals recorded on two adjacent contacts in the structure)
were preferred to monopolar signals. h2 values were over a
5 sec sliding window (1 sec step) for all possible pairs in both
hemispheres (200 sec of recordings in each auditory excerpt).
We computed a diagram of connectivity differences between
the mean h2 coefficient from both conditions for all combi-
nations of cortical structures. Both h2 (X- > Y) and h2 (Y- > X)
values have associated time delays t (X- > Y) and t (Y- > X)
between both signals of interest. By combining information
from the h2 and t measures, Dxy is expressed as:


Dxy ¼ 1=2 sgn h2 ðX/YÞ  h2 ðY/XÞ þ sgnðtðX/YÞ  tðY/XÞÞ
Dxy is comprised between 1 and 1, with positive Dxy
indicating that X(t) influences Y(t). To compare directionality
86 c o r t e x 6 0 ( 2 0 1 4 ) 8 2 e9 3
across conditions, we have determined the absolute value of
the difference of directionality between conditions. Signifi-
cant differences of directionality were plotted on the diagram
of h2 connectivity with arrows. The strength of the difference
was shown as a percentage of the distance between nodes in
the diagram (e.g., Dxy ¼ .1, i.e., Dxy will be presented as 10% of
the distance between X and Y on the diagram, starting from X).
In the case of directionality reversal between conditions, a
couple of arrows were plotted starting from the middle of the
distance between nodes, representing the direction and
strength of the relative condition.
To determine the statistical threshold in both calculations
(difference in h2 and Dxy) between the two conditions, we used
a permutation test (1000 permutations) based on the non-
parametric Wilcoxon rank sum test, with a Monte Carlo p-
value p ¼ .01, which was double side corrected for multiple
comparisons using the maximum cluster statistics described
by Maris & Oostenveld, 2007.
3. Results
CV is a non-musician but she used to listen to music, prefer-
ring variety rather than classical. She was not familiar with
the excerpts. She classified 37 out of 40 excerpts into the four
categories as performed by healthy subjects in the Bigand
et al.'s study (2005). Excerpts she liked had a mean rating of 8
for happy, 7.3 for sad, or 6.5 for peaceful, respectively. The
“angry” excerpts were unpleasant for the patient as she only
gave them a mean rating of 3.2. She classified these latter
excerpts in the right category, although she did not recognize
them as angry rather, she described them as aggressive and
that most of them evoked war for her.
3.1. Brain activation patterns
Cortical activation patterns can be observed on averaged
time-frequency energy maps computed for musical excerpts
as a function of emotion, and for pure tones. A significant
difference in frequency band activity from the different areas
of AC in both hemispheres is observed between pure tones
and music perception (Figs. 2 and 3).
The sequence of pure tones elicited a rhythmic theta ac-
tivity associated with an increase in low gamma band activity
(30e60 Hz), reflecting sharp synchronous evoked activity
rather than oscillatory processes (Fig. 2). This theta activation
was more reliably recorded from the right PAC (p < .0001) than
from the left PAC (p < .01) and was no longer observed either in
the SAC, BA 22 or AMY.
Listening to music entailed a significant enhancement in
the gamma band (40e100 Hz) compared to the silent periods
(p < .001) in PAC, SAC and BA 22, with a greater increase in
gamma (60e80 Hz) on the left side compared to the right side,
regardless of auditory areas or emotions. A marked decrease
of energy compared to the baseline below 20 Hz was observed
in lower frequency bands. In BA 22, the oscillatory pattern is
less visible, but the gamma activity was higher in the left than
right hemisphere (Fig. 3).
We conducted a 14 (contacts) 4 (emotion category) 4
(frequency bands) ANOVA with the power (energy frequency)
as the variable. We found that power is significantly different
according to the emotion [F (3, 4123) ¼ 17,5; MSE ¼ 2,19,
p < .0001], to the cortical area [F (13,4123) ¼ 20,98; MSE ¼ 2,62,
p < .0001], and to the frequency band [F (3,4123) ¼ 88,04;
MSE ¼ 11,04, p < .0001] with a significant interaction between
emotion X cortical areas [F (39,4123) ¼ 1,56, p < .01], emotion X
frequency bands [F (9,4123) ¼ 4,6, p < .0001] and cortical leads X
frequency bands [F (30,4123) ¼ 15,17, p < .0001]. Fig. 4 (top) il-
lustrates the interaction between cortical areas and emotion.
Angry excerpts stand out from all other emotions in the left
and right AMYe (p < .01 and p < .001 respectively) and in the
right SAC and BA 22 (p < .05). In addition, the arousal (contrast
between angry-happy vs sad-peaceful) is significantly modu-
lated in the right AMY (p < .0001), the right SAC and BA 22
(p < .05).
The contrasts between all the cortical areas and the fre-
quency bands show that the main distinction between the
emotions is found in the beta and gamma bands. An increase
in gamma band power and a decrease in beta band power
occurred more prominent in the left hemisphere, regardless
of cortical area, than in the right hemisphere (PAC and SAC,
p < .001, and p < .05 for BA 22). Post hoc analyses showed that
a significant difference between the emotions was observed
in the right AC. No significant difference in gamma band was
found in the left AC regardless of the emotion conveyed by
the musical excerpt (Fig. 4, bottom). Note that the left AMY is
removed from this latter analysis because of the recurrence
of epileptic spikes observed in the musical period, as well as
in the silent period, interfering with the gamma band
analysis.
A psychoacoustical model was used to test whether some
relevant features of the auditory signal may have been related
to SEEG signal. The basic psychoacoustic features of music
chosen for analysis were the level of dBA (loudness), zero-
crossing rate (ZCR), spectral centroid, roughness, pitch peri-
odicity flux (PP flux), and spectral roll-off. These features were
computed from the output of an auditory model by Leman
(2000). Although these features captured some of the behav-
ioral measures of arousal and valence in music (Lalitte &
Bigand, 2006), no one significant contribution of these fea-
tures were found with SEEG signal (p > .1). In addition, in order
to ascertain that the modulation is due to the emotion per se
and not to the different acoustical properties of each category
of musical emotion, we estimated the general rhythmic
pattern of both musical pieces and SEEG signals, calculating
the periodicities of these signals by means of the autocorre-
lation function (see Fig. 2 and text in supplemental material).
A 4 (emotion categories) 4 (areas) 2 (hemispheres) RM-
ANOVA, with pieces as the random variable, was conducted
on these correlation values. This analysis revealed a strong
main effect of hemisphere [F (1, 36) ¼ 24.63; MSE ¼ .1417;
p < .0005], indicating that the SEEG signals of the right hemi-
sphere were more related to the rhythmic pattern of the
musical signal than the left. The areas also had an effect on
the correlation values [F (3, 108) ¼ 7.2146; MSE ¼ .0182;
p < .0005]. More importantly, there was a main effect of the
emotional category of the pieces [F (3, 36) ¼ 5.28; MSE ¼ .0750;
p < .005], with sad pieces being associated with a very low
correlation value (p > .10) compared to the other three
emotions.
 c o r t e x 6 0 ( 2 0 1 4 ) 8 2 e9 3 87

Fig. 2 e Mean Time-Frequency plans (10 trials, 0e10 sec) during pure tone sequences. The pure tone bursts lasted 50 msec
(5 msec rise/fall time) and were delivered monaurally at four frequencies (.5, 1, 3, and 4 kHz) with an intensity set at 65 dB HL
at rate of one per second. The color coded energy from ¡5 mV blue to 5 mV red. The plans have been normalized by periods
of silence (10 trials 0e10 sec) during the resting state. The evoked activity (increase in theta-alpha and gamma power) is
constrained in PAC and SAC, whereas BA 22 and AMYG do not show a specific response to pure tones. Color Scale: ±5 mvlt.
88 c o r t e x 6 0 ( 2 0 1 4 ) 8 2 e9 3
Fig. 3 e Mean Time-Frequency plans (10 trials, ¡12 to
20 sec) during music listening (H:happy, S : sad, P: peaceful
and A: angry excerpts. The evoked and induced activities
spread in all PAC, SAC, BA22 and AMG. In PAC and SAC, an
increase in gamma power and a strong decrease in the
beta power are observed. Color Scale: ±5 mvlt.
3.2. Functional coupling between auditory areas and
AMY as a function of emotion
To probe interactions between AC and AMY in auditory pro-
cessing, we measured functional coupling between contacts
of each structure, computing the non-linear correlation coef-
ficient (h2) on intracerebral EEG signals. Fig. 5 shows a com-
parison of nonlinear correlation in periods with pure tones or
music versus resting state on the one hand, and according to
each emotion conveyed by the excerpts on the other hand
(ManneWhitneyeWilcoxon rank sum test thresholded at
p < .01 double sided).
Fig. 4 e Graph of interactions between emotions and
regions Top: a two way ANOVA (energy ¼ region *
emotion) shows significant differences between emotions
in the regions: L. AMY [F (3,4315) ¼ 4.21, p < .05], L. BA 22; [F
(3,4315) ¼ 3.73, p < .05), R. AMY; [F (3,4315) ¼ 6.21, p < .01],
R. SAC; [F (3,4315) ¼ 4.13, p < .05] and R. BA 22; [F
(3,4315) ¼ 6.32, p < .01]. Bottom: the same model of ANOVA
on data for gamma frequency band shows significant
differences between emotions in the regions: L. PAC [F
(3,986) ¼ 5.12, p < .01], R. PAC; [F (3,986) ¼ 6.16, p < .01] and
R. SAC; [F (3,986) ¼ 8.44, p < .001]. In purple, Tukey post-hoc
tests after a one way ANOVA (energy ¼ region) show
significant differences between sides for L. PAC versus R.
PAC (p < .001), L. SAC versus R. SAC (p < .001) and L. BA 22
versus R. BA 22 (p < .001).
Passive listening to pure tone sequences, as opposed to
attentive listening to music, revealed a striking difference in
the pattern of non-linear correlations between the auditory
areas and AMY. Pure tones elicit a significant widespread in-
crease of intra- and inter-hemispheric correlations between
PAC and SAC, as well as between the left AMY and all left
auditory areas and right SAC. Interestingly, the right AMY is
only correlated with the left PAC. A decrease of correlation
could also be observed between BA 22 and PAC and SAC.
Significant directionality, compared to the silent period, was
observed from right and left PAC towards SAC and BA22 and
from the right to the left PAC.
Music elicits a strong right lateralized network involving
increased correlation between the right AMY and PAC. Inter-
estingly, the right hemisphere auditory network (PACeSAC
and BA22) is interconnected, whereas the left hemisphere
auditory network shows a lower correlation compared to the
silent period. There was also no longer any involvement of the
left AMY. Analysis of the correlation according to the emotion
revealed the modulation of the functional connectivity be-
tween the AC and AMYe. An increased correlation between
right AMY and auditory areas for happy, peaceful and sad
 c o r t e x 6 0 ( 2 0 1 4 ) 8 2 e9 3 89

Fig. 5 e Diagrams of h2 connectivity (results of statistical comparing conditions) and directionality (p < .005). All the values of
connectivity are displayed in Table 1 (Supplemental material). The black arrows indicate the directionality of the
connectivity between structures expressed in ms. A e Connectivity patterns during Music (left) listening and pure tones
(right) hearing. Music listening shows a structured network involving the right AMY which is strongly connected to the PAC
structure. During Pure tones hearing, a larger network takes place where the left AMY displays a strong connectivity with
the left hemisphere auditory network. Right BA22 shows a strong disengagement in connectivity compared with resting
silence. B e Connectivity pattern during music listening, classified according to emotion (see comments in the text).

excerpts is observed. In contrast, angry excerpts led to a
widely distributed decreased correlation over auditory areas
on both sides and the left AMY. We failed to find any signifi-
cant difference in directionality during music listening
compared to a period of silence.
In a second step, we assessed whether emotion in music
emerged from a functional coupling between AC (3 areas) and
AMY: a 4 (emotions pieces) 4 (time periods) 4 (areas) 2
(hemisphere) ANOVA was performed, with the piece as the
random variable. The h2 defined the dependent measure.
90 c o r t e x 6 0 ( 2 0 1 4 ) 8 2 e9 3
Results for h2 were higher in the right than left hemisphere, F
(1,36) ¼ 97.18, MSE ¼ .0073363, p < .001, and linearly increased
in the right hemisphere when going from the coupling PAC-
AMY to the coupling BA22-AMY, as seen by a 2 hemisphere 3
areas interaction, F (2,72) ¼ 6.89. MSE ¼ .0014449, p < .002. The
effect of time period was expressed in a marginally significant
two way interaction between the time periods and the emo-
tions, F (9, 108) ¼ 1.92, MSE ¼ .0053992, p < .06. As illustrated in
Fig. 6, the coupling between cortical areas and the AMY
increased in musical excerpts classified as happy, and, to a
lesser extent, in excerpts classified as peaceful or sad, in
comparison to the preceding and following periods of silence.
The coupling considerably decreased for music classified as
angry. The contrast was strongest between angry and happy
music, with a highly significant difference in h2 values over
the 2 musical periods for these emotions F(1,36) ¼ 12.03,
MSE ¼ .076, p < .005 This demonstrates that the correlation/
decorrelation of the EEG signal recorded from auditory areas
and AMY marked the response to emotional music, with
strong decorrelation in the right hemisphere for angry music.
4. Discussion
In this study, we attempted to characterize the neural dy-
namic of emotion conveyed by musical stimuli. We analyzed
continuous brain activity patterns recorded from the AC and
AMY during music listening compared to a sequence of pure
tones. We used orchestral excerpts in an attentive listening
task that maintained the ecological validity of the listening
experience. The main results were: (i) an enhanced power of
gamma band oscillations in the AC evoked by music and not
by pure tones, which was associated with an increased spec-
tral complexity of sounds; (ii) the presence of asymmetry of
cortical tuning between the left and right AC, the right being
especially sensitive to emotion modulated by the variations of
the musical pieces; and (iii) the functional coupling between
the right AC and AMY, which is modulated by the correlation/
Fig. 6 e Functional coupling between AMY and auditory
areas as indexed by h2 as a function of the time periods
(pre-stimulation, two musical segments, and post-
stimulation) and the emotional content of the music.
decorrelation of the EEG signal and the self-evaluated
emotional valence.
4.1. Asymmetric emotional musical processing
The observation that gamma band activity is enhanced for
musical sequences compared to pure tone sequences is
consistent with evidence implicating gamma oscillations in
the processing of spectral complexity (Shahin, Roberts, Chau,
Trainor, & Miller, 2008; Shahin, Trainor, Roberts, Backer, &
Miller, 2010). An asymmetry in auditory cortical activity in
EEG rhythms was found during music listening, with the left
AC (preferentially PAC and SAC) showing stronger high fre-
quency fluctuations (high gamma 60e80 Hz) than the right
one, but being insensitive to emotions. In contrast, the right
AC was sensitive to emotions. This difference observed be-
tween the left and right auditory cortices could be related to
the intrinsic cortical oscillation frequencies which are
distinctive features of each sensory cortex; this tuning dif-
ference optimizes the sensitivity of the left hemisphere for
speech perception (high acoustic modulation) and the right
hemisphere for musical rhythms, musical instrument peri-
odicity or speech prosody (for review see, Giraud & Poeppel,
2012). In line with that reasoning, our findings showed that
the EEG modulation recorded from the right AC was more
related than the left AC to the rhythmic pattern of the musical
signal in relation to emotional processing. In contrast, we did
not observe any fluctuation in any EEG band in relation to the
emotions from the left AC, regardless of auditory area. It
should be noted that this hemispheric functional asymmetry
for music perception is present at birth. Perani et al. (2010)
showed that the fMRI BOLD signal in the right AC decreased
with altered music compared to original music, whilst no such
difference in BOLD response was observed in the left AC.
Beyond the right AC dominance, additional support for the
contribution of the right temporal lobe in music processing
comes from studies of patients with brain damage to the right
temporal lobe who have specific deficits in melody processing

geois-Chauvel, Peretz, Babai, Laguitton, & Chauvel, 1998;
(Lie
for a review, see Stewart, von Kriegstein, Warren, & Griffiths,
2006). Our analysis of frequency band dynamics across the
orchestral excerpts demonstrated a bilateral AMYr reactivity
as previously observed in functional imagery studies (Blood &
Zatorre, 2001; Koelsch, Fritz, Cramon D.Y., Muller, & Friederici,
2006; Mitterschiffthaler et al., 2007).
4.2. Functional coupling between auditory areas and the
AMY
Passive listening to a sequence of tone bursts elicited a
correlation between intra- et inter-hemispheric auditory
areas with a directionality from primary area to secondary
and associative areas, as observed in previous studies
demonstrating downstream propagation of elementary
acoustic stimuli (Gueguin, Le Bouquin-Jeannes, Faucon, &

geois-Chauvel, 2007; Kumar, Stephan, Warren, Friston, &
Lie
Griffiths, 2007). The functional coupling between-AC and
AMY evidences the anatomical connections between both
structures (LeDoux, 2007) but it takes place differently ac-
cording to the nature of the auditory stimuli. A specific
 c o r t e x 6 0 ( 2 0 1 4 ) 8 2 e9 3
correlation between the left AMY and the left AC is observed
during tone listening, which is no longer present during
music listening. This coupling could reflect pulse processing
elicited by the perception of roughly equally spaced and
salient tones (for review, see Grahn, 2012), driven by the
oscillatory properties of the left AC (for review, see Giraud &
Poeppel, 2012).
Interestingly, music listening drastically reduces the
functional correlation between all these areas, suggesting that
correlation/decorrelation between the AC and AMY builds up
a functional connection during the listening period, the
modulation of which is the neural signature of emotional
feeling. The role of such functional coupling has been studied
in the production of semiology in epileptic seizures. It has
been demonstrated that the correlation of intracerebral EEG
between certain cortical areas in certain frequencies is
therefore an electrophysiological counterpart of the emer-
gence of some clinical signs (Chauvel & McGonigal, 2014). For
instance, non-linear regression analysis (h2 as used in this
study) of depth-EEG signals showed the existence of transient
functional coupling between AMY and hippocampus and be-
tween hippocampus and rhinal cortex in “de
ja
 vu.”.
(Bartolomei et al., 2012). In addition, the mechanisms under-
lying visual percepts result from the phase locked synchrony
of neurons spatially distributed in the visual system in the
beta/gamma frequency range (Buzsa
ki & Draguhn, 2004;
Singer, 2004). In line with these empirical data, it is likely
that the induced emotion results from the functional coupling
between the AC and the AMY.
This can be emphasized by the absence of significant
directionality between structures (calculated from the h2), so
that a complex and non-stationary dialog is taking place be-
tween top-down structures while listening to music. Strik-
ingly, the emotional feeling is likely in relation to the coupling
found between the right AC and AMY. The changes observed
in iEEG signals in response to highly arousing and negative
valence pieces (usually categorized as angry) drastically
differed from the changes associated with all the other pieces.
One explanation could be that the patient did not listen as
attentively to angry excerpts as compared to the other musical
pieces. But passive listening to pure tones elicited a strong
correlation between AMY and AC. The distinction between
evoked feeling and conveyed emotions has been a matter of
long debate. The differences between felt and perceived
emotion are difficult to disentangle but most often the nega-
tive emotions, although not reported to be felt in response to
music, are reported to be perceived as expressive properties of
the music (Zentner et al., 2008). But in this study, angry ex-
cerpts (even though the label of angry is not the most appro-
priate) induced an emotive state different from the others
since the decorrelation found between AC and AMY might
depend on the displeasure experienced by the patient
listening to music she did not like. The response could
therefore be related to emotions conveyed by music as well as
evoked feelings. There is a large body of evidence illustrating
how sound can cause functional modifications in the limbic
system which, in turn, could modulate auditory plasticity
(Kumar, von Kriegstein, Friston, & Griffiths, 2012; for review,
see Kraus & Canlon, 2012). The connectivity analysis using the
dynamic causal modeling revealed that the sound is first
91
processed in the AC and the AMY secondarily modulates the
AC as a function of valence.
4.3. Limitations of the study
We must acknowledge some limitations to our study. It is a
single case study and as such, these results should be inter-
preted with caution. Furthermore, it is largely admitted that a
single piece of music would not induce the same emotion in
all subjects, the subjective experience, i.e., liking or disliking a
specific piece of music, depends on the individual's biography,
musical exposure and preferences (Brattico & Jacobsen, 2009;
McDermott, Lehr & Osenham, 2010; Nieminen, Istok,
Brattico, Tervaniemi &Huotilainen, 2011; Shahin et al., 2010).
In addition, recent studies analyzing individual brain-
frequency responses either to musical excerpts chosen by
the experimenter (Lin, Duann, Chen, & Jung, 2010) or to self-
selected music (Holler et al., 2012) showed inter-individual
differences in emotion-related spectral changes while
listening to the music.
In conclusion, this case study points out that: (i) the audi-
toryelimbic interactions result from the formation of oscilla-
tory networks binding the activities of the network nodes,
promoting the emergence of a feeling and (ii) that the music is
related to both sensory and cognitive processes preferentially
relayed in the right hemisphere.
Acknowledgments
We thank Pr J Regis (Neurosurgery Department, Marseille) for
stereotactic placement of the electrodes, D Scho€ n and V Jirsa
for fruitful discussions and comments on the manuscript and
R Mc Cully for editing english. We thank Patrick Marquis for
dedicated help in data acquisition. We wish to thank the pa-
tient C.V for her active cooperation. The work was supported
by a grant (ACI Neurosciences integrative et computationelles)
from the French Government.
Supplementary data
Supplementary data related to this article can be found at
http://dx.doi.org/10.1016/j.cortex.2014.06.002.
references
Amaral, D. G., & Price, J. L. (1985). Amygdalo-cortical projections
in the monkey (Macaca fascicularis). The Journal of Comparative
Neurology, 230, 465e496.
Bartolomei, F., Barbeau, E. J., Nguyen, T., McGonigal, A., Re
gis, J.,
Chauvel, P., et al. (2012). Rhinalhippocampal interactions
during de
ja
 vu. Clinical Neurophysiology, 123, 489e495.
Baumgartner, T., Esslen, M., & Ja € ncke, L. (2006). From emotion
perception to emotion experience: emotions evoked by
pictures and classical music. International Journal of
Psychophysiology, 60(1), 34e43.
Behne, N., Scheich, H., & Brechmann, A. (2005). Contralateral
white noise selectively changes right human auditory cortex
92 c o r t e x 6 0 ( 2 0 1 4 ) 8 2 e9 3
activity caused by a FM-Direction task. Journal of
Neurophysiology, 931, 414e423.
Bigand, E., Filipic, S., & Lalitte, P. (2005). The time course of
emotional responses to music. Annals of the New York Academy
of Sciences, 1060, 429e437.
Blood, A. J., Zatorre, R. J., Bermudez, P., & Evans, A. C. (1999).
Emotional responses to pleasant and unpleasant music
correlate with activity in paralimbic brain regions. Nature
Neuroscience, 2, 382e387.
Blood, A., & Zatorre, R. J. (2001). Intensely pleasurable responses
to music correlate with activity in brain regions implicated in
reward and emotion. Proceedings of the National Academy of
Sciences of the United States of America, 98, 11818e11823.
Brattico, E., Tervaniemi, M., Na €a€ ta
€ nen, R., & Peretz, I. (2006).
Musical scale properties are automatically processed in the
human auditory cortex. Brain Research, 1117, 162e174.
Brattico, E., & Jacobsen, T. (2009). Subjective appraisal of music:
Neuroimaging evidence. Annals of the New York Academy of
Sciences, 1169, 308e317.
Brechmann, A., & Scheich, H. (2005). Hemispheric shifts of sound
representation in auditory cortex with conceptual listening.
Cerebral Cortex, 155, 578e587.
Buzsa
ki, G., & Draguhn, A. (2004). Neuronal oscillations in cortical
networks. Science, 304, 1926e1929.
Chauvel, P., & McGonigal, A. (2014). Emergence of semiology in
epileptic seizures. Epilepsy and Behavior. in press [Epub ahead
of print].
Dalla Bella, S., Peretz, I., Rousseau, L., Gosselin, N., Ayotte, J., &
Lavoie, A. (2001). Development of the happy-sad distinction in
music appreciation. Does tempo emerge earlier than mode?
Annals of the New York Academy of Sciences, 930, 436e438.
Eldar, E., Ganor, O., Admon, R., Bleich, A., & Hendler, T. (2007).
Feeling the real world limbic response to music depends on
related content. Cerebral Cortex, 17, 2828e2840.
Filipic, S., Tillmann, B., & Bigand, E. (2010). Judging familiarity and
emotion from very brief musical excerpts. Psychonomic Bulletin
& Review, 17(3), 335e341.
Fritz, T., Jentschke, S., Gosselin, N., Sammler, D., Peretz, I.,
Turner, R., et al. (2009). Universal recognition of three basic
emotions in music. Current Biology, 19, 1e4.
Giraud, A. L., & Poeppel, D. (2012). Cortical oscillations and speech
processing: emerging computational principles and
operations. Nature Neuroscence, 15(4), 511e517.
Gosselin, N., Samson, S., Adolphs, R., Noulhiane, M., Roy, M.,
Hasboun, D., et al. (2006). Emotional responses to unpleasant
music correlates with damage to the parahippocampal cortex.
Brain, 129, 2585e2592.
Grandjean, D., Sander, D., Pourtois, G., Schwartz, S., Seghier, M. L.,
Scherer, K. R., et al. (2005). The voices of wrath: brain
responses to angry prosody in meaningless speech. Nature
Neuroscience, 8(2), 145e146.
Grahn, J. A. (2012). Neural mechanisms of rhythm perception:
current findings and future perspectives. Topics in Cognitive
Science, 4, 585e606.
Gueguin, M., Le Bouquin-Jeannes, R., Faucon, G., & Lie
geois-
Chauvel, C. (2007). Evidence of functional connectivity
between auditory cortical areas revealed by amplitude
modulation sound processing. Cerebral Cortex, 17(2), 304e313.
Hoistad, M., & Barbas, H. (2008). Sequence of information processing
for emotions through pathways linking temporal and insular
cortices with the amygdale. NeuroImage, 40, 1016e1033.
Ho€ ller, Y., Thomschewski, A., Schmid, E. V., Ho € ller, P., Crone, J. S.,
& Trinka, E. (2012). Individual brain-frequency responses to
self-selected music. International Journal of Psychophysiology,
86(3), 206e213.
Johnsrude, I. S., Penhune, V. B., & Zatorre, R. J. (2000). Functional
specificity in right human auditory cortex for perceiving pitch
direction. Brain, 123, 155e163.
Juslin, P. N., & Va€ stfja
€ ll, D. (2008). Emotional responses to music:
the need to consider underlying mechanisms. The Behavioral
and Brain Sciences, 31, 559e621.
Juslin, P. N. (2013). From everyday emotions to aesthetic
emotions: towards a unified theory of musical emotions.
Physics of Life Reviews, 10(3), 235e266.
Khalfa, S., Schon, D., Anton, J. L., & Lie
geois-Chauvel, C. (2005).
Brain regions involved in the recognition of happiness and
sadness in music. NeuroReport, 16(18), 1981e1984.
Khalfa, S., Guye, M., Peretz, I., Chapon, F., Girard, N., Chauvel, P.,
et al. (2008a). Evidence of lateralized anteromedial temporal
structures involvement in musical emotion processing.
Neuropsychologia, 4610, 2485e2493.
Khalfa, S., Delbe
, C., Bigand, E., Reynaud, E., Chauvel, P., &
Liegeois-Chauvel, C. (2008b). Positive and negative music
recognition reveals a specialization of mesio-temporal
structures in epileptic patients. Music Perception, 25(4),
295e302.
Koelsch, S., Fritz, T., V. Cramon, D. Y., Muller, K., &
Friederici, A. D. (2006). Investigating emotion with music. A
fMRI study. Human Brain Mapping, 27, 239e250.
Kraus, K. S., & Canlon, B. (2012). Neuronal connectivity and
interactions between the auditory and limbic systems effects
of noise and tinnitus. Hearing Research, 288, 34e46.
Krumhansl, C. L. (1997). An exploratory study of musical
emotions and psychophysiology. Canadian Journal of
Experimental Psychology, 51, 336e353.
Kumar, S., Stephan, K. E., Warren, J. D., Friston, K. J., &
Griffiths, T. D. (2007). Hierarchical processing of
auditory objects in humans. PLos Computational Biology,
3(6), e 100.
Kumar, S., von Kriegstein, K., Friston, K., & Griffiths, T. D. (2012).
Features versus feelings: dissociable representations of the
acoustic features and valence of aversive sounds. The Journal of
Neuroscience, 32(41), 14184e14192.
Labbe, C., & Grandjean, D. (2014). Musical emotions predicted by
feelings of entrainment. Music Perception. in press.
Lalitte, P., & Bigand, E. (2006). Music in the moment? Revisiting
the effect of large scale structures. Perceptual and Motor Skills,
103, 811e828.
LeDoux, J. (2007). The amygdala. Current Biology, 17, 868e874.
Leman, M. (2000). An auditory model of the role of short term
memory in probe tone ratings. Music Perception, 17(4), 481e509.
Levitin, D. J., & Menon, V. (2003). Musical structure is processed in
‘‘language’’ areas of the brain: a possible role for Brodmann
area 47 in temporal coherence. NeuroImage, 20, 2142e2152.

geois-Chauvel, C., Musolino, A., & Chauvel, P. (1991).
Lie
Localization of the primary auditory area in man. Brain, 114,
139e151.

geois-Chauvel, C., Musolino, A., Badier, J. M., Marquis, P., &
Lie
Chauvel, P. (1994). Evoked potentials recorded from the
auditory cortex in man: evaluation and topography of the
middle latency components. Electroencephalography and Clinical
Neurophysiology, 923, 204e214.

geois-Chauvel, C., Peretz, I., Babai, M., Laguitton, V., &
Lie
Chauvel, P. (1998). Contribution of the different temporal
cortical structures to music processing. Brain, 121, 1853e1867.

geois-Chauvel, C., de Graaf, J. B., Laguitton, V., & Chauvel, P.
Lie
(1999). Specialization of left auditory cortex for speech
perception in man depends on temporal coding. Cerebral
Cortex, 95, 484e496.

geois-Chauvel, C., Giraud, K., Badier, J. M., Marquis, P., &
Lie
Chauvel, P. (2001). Intracerebral evoked potentials in pitch
perception reveal a functional asymmetry of the human
auditory cortex. Annals of the New York Academy of Sciences, 930,
117e132.
Lin, Y. P., Duann, J. R., Chen, J. H., & Jung, T. P. (2010).
Electroencephalographic dynamics of musical emotion
 c o r t e x 6 0 ( 2 0 1 4 ) 8 2 e9 3
perception revealed by independent spectral components.
NeuroReport, 21(6), 410e415.
Lopes da Silva, F., Pijn, J. P., & Boeijinga, P. (1989).
Interdependence of EEG signals: linear vs nonlinear
associations and the significance of time delays and phase
shifts. Brain Topography, 21(2), 9e18.
McDermott, J., Lehr, A., & Oxenham, A. (2010). Individual
differences reveal the basis of consonance. Current Biology, 20,
1035e1041.
Maris, E., & Oostenveld, R. (2007). Non parametric statistical
testing of EEG and MEG Data. Journal of Neuroscience Methods,
164(1), 177e190.
Menon, V., & Levitin, D. J. (2005). The rewards of music listening:
response and physiological connectivity of the mesolimbic
system. NeuroImage, 28, 175e184.
Mitterschiffthaler, M. T., Fu, C. H. Y., Dalton, J. A., Andrew, C. M.,
& Williams, S. C. R. (2007). A functional MRI study of happy
and sad affective states induced by classical music. Human
Brain Mapping, 21, 1150e1162.
Nieminen, S., Isto
k, E., Brattico, E., Tervaniemi, M., &
Huotilainen, M. (2011). The development of aesthetic
responses to music and their underlying neural and
psychological mechanisms. Cortex, 47, 1138e1146.
Patterson, R. D., Uppenkamp, S., Johnsrude, I. S., & Griffiths, T. D.
(2002). The processing of temporal pitch and melody
information in auditory cortex. Neuron, 36, 767e776.
Perani, D., Saccuman, M. C., Scifo, P., Spada, D., Andreolli, G.,
Rovelli, R., et al. (2010). Functional specializations for music
processing in the human newborn brain. Proceedings of the
National Academy of Sciences of the United States of America,
107(10), 4758e4763.
Pereira, C. S., Teixeira, J., Figueiredo, P., Joao, X., Castro, S. L., &
Brattico, E. (2011). Music and emotions in the brain: familiarity
matters. PLos One, 11(6), 27241.
Peretz, I., & Hebert, S. (2000). Toward a biological account of
music experience. Brain Cognition, 42, 131e134.
Peretz, I., & Zatorre, R. (2005). Brain organization for music
processing. Annual Review of Psychology, 56, 89e114.
Plichta, M. M., Gerdes, A. B. M., ALpers, G. W., Harnisch, W.,
Brill, S., Wieser, M. J., et al. (2011). Auditory cortex activation is
modulated by emotion: a functional near-infrered
spectroscopy (fNIRS) study. NeuroImage, 55, 1200e1207.
93
Price, J. L. (2003). Comparative aspects of amygdale connectivity.
Annals of the New York Academy of Sciences, 985, 50e58.
Samson, S., & Zatorre, R. J. (1988). Melodic and harmonic
discrimination following unilateral cerebral excision. Brain
Cognition, 7, 348e360.
Schmithorst, V. J., & Holland, S. K. (2003). The effect of musical
training on music processing: a functional magnetic
resonance imaging study in humans. Neuroscience Letters, 348,
65e68.
Shahin, A. J., Roberts, L. E., Chau, W., Trainor, L. J., & Miller, L. M.
(2008). Music training leads to the development of timbre-
specific gamma band activity. NeuroImage, 411, 113e122.
Shahin, A., Trainor, L., Roberts, L., Backer, K., & Miller, L. (2010).
Development of auditory phase-locked activity for music
sounds. Journal of Neurophysiology., 103, 218e229.
Singer, W. (2004). Synchrony, oscillations and relational codes.
Visual Neuroscience, 2, 1665e1681.
Stewart, L., von Kriegstein, K., Warren, J. D., & Griffiths, T. D.
(2006). Music and the brain: disorders of musical listening.
Brain, 129, 2533e2553.
Trochidis, K., & Bigand, E. (2013). Investigation of the effect of
mode and tempo on emotional responses to music using EEG
power asymmetry. Journal of Psychophysiology, 27, 142e148.
Warrier, C. M., & Zatorre, R. J. (2004). Right temporal cortex is
critical for utilization of melodic contextual cues in a pitch
constancy task. Brain, 127, 1616e1625.
Wendling, F., Bartolomei, F., Bellanger, J. J., & Chauvel, P. (2001).
Interpretation of interdependencies in epileptic signals using
a macroscopic physiological model of the EEG. Clinical
Neurophysiology, 112, 1201e1218.
Wendling, F., Bartolomei, F., Bellanger, J. J., Bourien, J., &
Chauvel, P. (2003). Epileptic fast intracerebral EEG activity:
evidence for spatial decorrelation at seizure onset. Brain, 126,
1449e1459.
Zatorre, R. J., & Belin, P. (2001). Spectral and temporal processing
in human auditory cortex. Cerebral Cortex, 1110, 946e953.
Zatorre, R. J., Belin, P., & Penhune, V. B. (2002). Structure and
function of auditory cortex: music and speech. Trends in
Cognitive Sciences, 61, 37e46.
Zentner, M., Grandjean, D., & Scherer, K. R. (2008). Emotions
evoked by the sound of music: characterization, classification
and measurement. Emotion, 8(4), 494e521.