Edited by: Lars Schwabe, University of Hamburg, Germany
Reviewed by: Phillip R. Zoladz, Ohio Northern University, USA; Mathias Weymar, University of Greifswald, Germany
*Correspondence: Louise Bonnet, Stroke Unit, Department of Neurology, Besancon University Hospital, 3, bd Alexandre Fleming, 25030 Besancon, France
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
The specific role of the amygdala remains controversial even though the development of functional imaging techniques has established its implication in the emotional process. The aim of this study was to highlight the sensitivity of the amygdala to emotional intensity (arousal). We conducted an analysis of the modulation of amygdala activation according to variation in emotional intensity via an fMRI event-related protocol. Monitoring of electrodermal activity, a marker of psychophysiological emotional perception and a reflection of the activation of the autonomic nervous system, was carried out concurrently. Eighteen subjects (10 men; aged from 22 to 29 years) looked at emotionally positive photographs. We demonstrated that the left and right amygdalae were sensitive to changes in emotional intensity, activating more in response to stimuli with higher intensity. Furthermore, electrodermal responses were more frequent for the most intense stimuli, demonstrating the concomitant activation of the autonomic nervous system. These results highlight the sensitivity of the amygdala to the intensity of positively valenced visual stimuli, and in conjunction with results in the literature on negative emotions, reinforce the role of the amygdala in the perception of intensity.
One of the main challenges of neuroscience in the field of emotions is the modeling of anatomical structure and its functional response underlying emotional experience. The two main models of emotion, categorical, and dimensional, are each underpinned by neural and psychophysiological patterns consistent with and validated by numerous studies (Anderson et al.,
The amygdala is an anatomical and functional crossroads in the emotional process of which the role has been studied by both the categorical and dimensional models of emotions. Initially, the amygdala was considered to be “the organ of fear” according to a categorical model (LeDoux,
The dimensional model now seems more appropriate in studying the role of the amygdala in emotion. The challenge was then to decide if the amygdala encodes information about valence or intensity of emotion. First, it has been shown that the amygdala is activated in correlation with the intensity of negative stimuli (Canli et al.,
In regard to visual emotional stimulation, the most widely used stimuli are images from the International Affective Picture System (IAPS). Each of these images is provided with a normalized value of valence (
The subject's perception of the two dimensions of valence and intensity in emotional feeling can be evaluated by self-scoring. Furthermore, emotions can be objectively evaluated by monitoring the autonomic nervous system (Lang et al.,
The aim of our study was to demonstrate the role of the amygdala in the perception of the intensity of pleasant emotions when using positively-valenced images without using negative emotional stimuli (unpleasant). In the MRI scanner, we simultaneously recorded electrodermal activity (EDA) in order to indicate the emotional intensity perceived by the subject. In order to control the effect of our stimuli, we replicated our protocol of emotional stimulation outside the scanner 1 month later to evaluate the effect of our stimuli, without the effect of the scanner. Our hypothesis is that increasing the intensity of positive visual emotional stimuli increases amygdala activation. This result should complement knowledge regarding negative as well as positive emotions in other domains (using olfactory, auditory, gustatory or semantic stimuli). We could then consider the fact that the human amygdala encodes the intensity of emotional stimuli whether positive or negative.
Eighteen healthy participants (10 men, all right handed, mean age 25 years, range 22–29) with normal or corrected-to-normal vision, and no history of neurological and psychiatric disorders, participated in this study. Written informed consent, as well as a safety-screening questionnaire to undergo magnetic resonance imaging (MRI), was obtained from each participant. The study received the approval of an Institutional Review Board (CPP
Stimuli were 75 positively valenced (IAPS ratings >4.6; mean = 5.90 ± 0.49) colored visual stimuli selected from the IAPS (Lang et al.,
Lizard | 1121 | 5, 79 | 4, 83 |
Leopard | 1310 | 4, 6 | 6 |
Gannet | 1450 | 6, 37 | 2, 83 |
Jaguar | 1650 | 6, 65 | 6, 23 |
Octopus | 1947 | 5, 85 | 4, 35 |
Woman | 2025 | 5, 78 | 4, 3 |
Cheerleaders | 2034 | 5, 9 | 4, 93 |
NeuWoman | 2038 | 5, 09 | 2, 94 |
Girl | 2320 | 6, 17 | 2, 9 |
ThreeMen | 2370 | 7, 14 | 2, 9 |
Girl | 2411 | 5, 07 | 2, 86 |
ManW/Dog | 2521 | 5, 78 | 4, 1 |
Chess | 2580 | 5, 71 | 2, 79 |
Beer | 2600 | 5, 84 | 4, 16 |
Dance | 2606 | 5, 92 | 4, 78 |
Dancer | 2616 | 5, 97 | 4, 96 |
Woman | 2620 | 5, 93 | 2, 72 |
Shadow | 2880 | 5, 18 | 2, 96 |
Gold | 3005, 2 | 5, 98 | 4, 84 |
EroticFemale | 4232 | 5, 95 | 6, 28 |
EroticMale | 4490 | 6, 27 | 6, 06 |
EroticMale | 4503 | 6 | 4, 93 |
EroticMale | 4531 | 5, 81 | 4, 28 |
EroticMale | 4538 | 5, 91 | 4, 65 |
EroticCouple | 4604 | 5, 98 | 6, 09 |
EroticCouple | 4651 | 6, 32 | 6, 34 |
EroticCouple | 4669 | 5, 97 | 6, 11 |
EroticCouple | 4692 | 5, 87 | 6, 39 |
EroticCouple | 4693 | 6, 16 | 6, 57 |
EroticCouple | 4697 | 6, 22 | 6, 62 |
Flower | 5000 | 7, 08 | 2, 67 |
Flower | 5020 | 6, 32 | 2, 63 |
Flower | 5030 | 6, 51 | 2, 74 |
Boat | 5390 | 5, 59 | 2, 88 |
Cockpit | 5455 | 5, 79 | 4, 56 |
Mushroom | 5520 | 5, 33 | 2, 95 |
Mushroom | 5530 | 5, 38 | 2, 87 |
HangGlider | 5626 | 6, 71 | 6, 1 |
Cave | 5661 | 5, 96 | 4, 15 |
Farmland | 5720 | 6, 31 | 2, 79 |
Grain | 5726 | 6, 23 | 2, 84 |
Flowers | 5731 | 5, 39 | 2, 74 |
Leaves | 5800 | 6, 36 | 2, 51 |
Desert | 5900 | 5, 93 | 4, 38 |
Volcano | 5920 | 5, 16 | 6, 23 |
Lightning | 5950 | 5, 99 | 6, 79 |
Picnic Table | 7026 | 5, 38 | 2, 63 |
Fork | 7080 | 5, 27 | 2, 32 |
Headlight | 7095 | 5, 99 | 4, 21 |
Fire Hydrant | 7100 | 5, 24 | 2, 89 |
Bus | 7140 | 5, 5 | 2, 92 |
Teeth | 7195 | 6, 02 | 4, 54 |
Scarves | 7205 | 5, 56 | 2, 93 |
Pizza | 7351 | 5, 82 | 4, 25 |
Sushi | 7477 | 6, 12 | 4, 82 |
Window | 7490 | 5, 52 | 2, 42 |
Street | 7496 | 5, 92 | 4, 84 |
Card Dealer | 7503 | 5, 77 | 4, 21 |
Stairs | 7504 | 5, 67 | 4, 25 |
Skyscraper | 7510 | 6, 05 | 4, 52 |
Jet | 7620 | 5, 78 | 4, 92 |
Skyscraper | 7640 | 5 | 6, 03 |
City | 7650 | 6, 62 | 6, 15 |
Violin | 7900 | 6, 5 | 2, 6 |
Hiker | 8158 | 6, 53 | 6, 49 |
Rock Climber | 8160 | 5, 07 | 6, 97 |
Cliffdiver | 8178 | 6, 5 | 6, 82 |
Bungee | 8179 | 6, 48 | 6, 99 |
Ice Climber | 8191 | 6, 07 | 6, 19 |
Volcano Skier | 8192 | 5, 52 | 6, 03 |
Surfers | 8206 | 6, 43 | 6, 41 |
Motorcycle | 8251 | 6, 16 | 6, 05 |
Wingwalker | 8341 | 6, 25 | 6, 4 |
Biking/train | 8475 | 4, 85 | 6, 52 |
Woman | 8620 | 6, 04 | 4, 6 |
This study was composed of two sessions. In the first session, we used functional magnetic resonance brain imaging (fMRI) to examine regional brain activity during a spontaneous emotion reactivity task. Our marker of autonomic activation was EDA, monitored in the scanner. We chose a passive task in order to avoid potential inhibition of amygdala activity by the prefrontal cortex because task instructions involving a form of attentional processing reduce the likelihood of amygdala activation compared to the passive processing of emotional stimuli (Hariri et al.,
In the second phase, 1 month later, the same participants performed the same paradigm, but outside the scanner. After this second session, all stimuli were presented in a second pass where subjects rated how arousing they had experienced each stimulus on a scale ranging from non-arousing (intensity value of 1) to arousing (intensity value of 9) on a paper-and-pencil version of the self-assessment-manikin (Bradley and Lang,
Skin conductance was recorded using an MP-150 psychophysiological monitoring system (BioPac Systems, Santa Barbara, CA). We used Ag/AgCl electrodes filled with isotonic NaCl unibase electrolytes attached to the volar surface of the second phalanx of the second and third fingers of the left hand (non-dominant hand) (Fowles et al.,
The MP-150 module performed analog-to-digital conversion of the amplified signals and passed data to a computer running Acknowledge 4.2 software (BioPac Systems, Santa Barbara, CA) for analysis of waveforms.
The SCR data were processed using high-pass filter and smoothing to remove scanner-induced artifacts. The tonic component was then extracted from the phasic component, in order to suppress the effect of the precedent stimulus on the SCR of the next one (Lim et al.,
For each stimulus we calculated the frequency of SCRs (percentage of SCR among the 18 subjects), the magnitude (mean value computed across all stimulus presentations including those without a measurable response) and the amplitude (mean value computed across only those trials on which a measurable response occurred) of SCRs across all the subjects (Dawson et al.,
Correlations between both sessions (inside and outside the scanner) were calculated for frequencies as well as magnitudes.
Imaging data was collected at Besancon University Hospital using a 3-Tesla (General Electric Healthcare Signa H.D. Milwaukee, WI, USA) MR system with a standard 40 mT/m gradient using blood–oxygen level-dependent (BOLD) fMRI. Foam cushions were used to minimize head movements within the coil. Functional MRI runs were acquired parallel to the anterior-posterior commissural line, covering the entire cerebrum using an echo planar imaging (EPI) sequence: echo time (TE) = 35 ms, flip angle (FA) = 90°, matrix size = 128 * 128, field of view (FOV) = 256 mm, slice thickness = 4.5 mm, 30 slices and repetition time (TR) = 2500 ms. Before the first run, a high-resolution, T1-weighted, three-dimensional data set encompassing the whole brain was acquired to provide detailed anatomy (G.E. Fast Spoiled Gradient Recalled Echo sequence, matrix size = 256 * 256, FOV = 256 mm2, 134 slices, slice thickness = 1 mm, no gap, total scan time = 2 min 56 s).
Image time-series analysis was performed using BrainVoyager QX 2.4 (Brain Innovation, Maastricht, The Netherlands) (Goebel et al.,
The ratings of intensity and valence were normally distributed, according to a Shapiro-Wilk test, but there was no equality of variances across the means of intensity of our three groups (Bartlett test). We performed a Kruskall-Wallis test to search for differences between our three groups using the means of intensity (subjects' ratings) and valence (norms from the IAPS) for each stimulus. When a difference was detected, we performed a Student test to search for differences between each group of stimuli.
SCRs were not normally distributed (Shapiro-Wilk test). We used non-parametric tests. The differences in SCRs across the three groups of stimuli were assessed using a Kruskall-Wallis test and when a difference was found, comparison between groups was made by a Welch test. The differences in SCRs across the two sessions were assessed using a Wilcoxon test. Tests were undertaken for frequency of SCRs, magnitude, and amplitude (microsiemens).
Ratings made by our subjects were strongly correlated to IAPS norms of intensity (
According to these three new groups, mean luminance for the group of low intensity (respectively medium intensity, high intensity) was 87.3 (respectively 86.4, 89.9). No statistical difference was observed between groups neither for this parameter, nor for the contrast.
The global fMRI analysis comparing the viewing of stimuli and rest state showed activation in the right and left anterior cingulate gyri, right, and left superior medial frontal gyri, right, and left inferior frontal gyri, right, and left posterior orbital gyri, right, and left anterior part of insula, right, and left occipital gyri, right, and left thalami (medial thalamus and pulvinar), right and left colliculi, right, and left amygdalae, right, and left parahippocampal gyri, and the vermis. These areas are known to be involved in the processing of visual emotional information (Sabatinelli et al.,
The fMRI analysis between the three groups of stimuli (according to subjects' intensity ratings) showed activations in the right and left amygdalae, the right orbital gyrus, the right pulvinar, the right, and left medial thalami, the anterior part of the insula, the right, and left colliculi, and the hypothalamus (see Tables
Left amygdala | × | × |
Right amygdala | × | × |
Right orbital gyrus | × | × |
Right dorsolateral prefrontal cortex (inferior frontal gyrus) | × | Right dorsolateral prefrontal cortex (inferior frontal gyrus) |
Right medial frontal gyrus | Right medial frontal gyrus | × |
Left medial frontal gyrus | Left medial frontal gyrus | × |
Right fusiform gyrus | Right fusiform gyrus | × |
Left parahippocampal gyrus | ||
Right parahippocampal gyrus | × | Right parahippocampal gyrus |
Right temporo-occipital junction | × | Right temporo-occipital junction |
Left temporo-occipital junction | × | Left temporo-occipital junction |
Left parieto-occipital junction | × | × |
Right posterior cingulate | × | × |
Left posterior cingulate | × | Left posterior cingulate |
Left anterior insula | ||
Right and left colliculi | × | × |
Right pulvinar | × | × |
Left medial thalamus | × | Left medial thalamus |
Right medial thalamus | × | × |
Right hypothalamus | × | × |
Vermis | × | × |
Left amygdala | −22 | −8 | −9 | 5.134 | 777 | |
Right amygdala | 17 | −2 | −12 | 5.130 | 750 | |
Vermis | −4 | −59 | −43 | 6.081 | 923 | |
Right fusiform gyrus | 20 | 32 | −41 | −18 | 6.739 | 3139 |
Right orbital gyrus | 47 | 29 | 13 | −18 | 7.477 | 1898 |
Right temporo−occipital junction | 37 | 38 | −59 | 9 | 6.445 | 5821 |
Left temporo−occipital junction | 37 | −53 | −71 | 3 | 6.943 | 3838 |
Right parahippocampal gyrus | 28 | 23 | −29 | −9 | 4.910 | 1698 |
Left parahippocampal gyrus | 28 | −10 | −35 | −9 | 5.923 | 1907 |
Right colliculus | 6 | −32 | −2 | 4.848 | 112 | |
Left colliculus | −5 | −35 | −7 | 5.126 | 209 | |
Right pulvinar | 11 | −32 | 3 | 5.291 | 1059 | |
Left parieto−occipital junction | 19 | −16 | −59 | 3 | 4.788 | 377 |
Right posterior cingulate | 23 | 2 | −56 | 18 | 5.473 | 1579 |
Left posterior cingulate | 23 | −2 | −59 | 21 | 5.007 | 534 |
Left medial thalamus | −4 | −14 | −3 | 7.386 | 1259 | |
Right medial thalamus | 0 | −14 | −3 | 5.956 | 556 | |
Right dorsolateral prefrontal cortex (inferior frontal gyrus) | 9 | 38 | 10 | 27 | 6.032 | 1926 |
Left anterior insula | −34 | 28 | 0 | 4.119 | 68 | |
Right hypothalamus | 13 | −2 | −12 | 4.755 | 158 | |
Left medial frontal gyrus | 10 | −1 | 55 | 9 | 4.766 | 979 |
Regarding the main objective of the study, we found a stronger activation of the right and left amygdalae when subjects visualized stimuli of stronger intensity (Group 3) compared to stimuli from the group of lower intensity (Group 1). The left and right amygdalae were activated by viewing emotional stimuli, as demonstrated by the analysis of the differences of cerebral activation between no stimulation (black cross) and the viewing of pictures. However, amygdala activation was stronger when subjects viewed stimuli with strong intensity compared to stimuli with low intensity, with the same valence (neutral or positive). There was no significant difference with corrected statistics in intermediary comparison (between Groups 1 and 2 and between Groups 2 and 3) for the amygdalae. These results are illustrated on Figure
During the fMRI session, 16 subjects presented SCRs and 2 subjects (2 women) presented no SCRs. The magnitude and frequency of SCRs were greater for the stimuli in Group 3, rated by subjects as the most intense (
During the second phase, 17 subjects presented SCRs and 1 subject showed no SCRs (a woman who did not present any SCRs in the first phase). The frequency of SCRs for Group 3 was significantly higher than those of Groups 1 and 2 (
Frequency, magnitude, and amplitude of the SCRs were higher during the second pass, outside the MRI scanner, compared to the first pass in the MRI scanner (respectively: frequency = 14 vs. 27%; magnitude = 0.01 vs. 0.04; amplitude = 0.08 vs. 0.12;
Frequencies of electrodermal responses were significantly correlated between sessions 1 and 2 (
We have shown that the amygdala is sensitive to the emotional intensity of positive stimuli. This result is consistent with studies in primates (Belova et al.,
Our study did not reveal any specific amygdala lateralization. Various models of amygdala lateralization have been proposed, with more recent models showing that amygdala lateralization may be linked to the temporal dynamics of information processing (Wright et al.,
Our study showed activation of the medial thalamus and the posterior orbital and medial prefrontal cortices, parallel to the increase in the intensity of the stimuli. Concordantly with our study, a coding of emotional intensity by the medial thalamus and the medial prefrontal cortex was demonstrated (Anders et al.,
We have demonstrated that the right pulvinar and the left and right colliculi are sensitive to emotional intensity. Data for this circuit are controversial. For some, it may be responsible for the unconscious processing of visual emotional stimuli, while for others, its role, like that of the amygdala, may be to coordinate the neocortical circuits in the treatment of the relevance of visual emotional information (Buchsbaum et al.,
Activation of the hypothalamus according to emotional intensity is consistent with amygdala activation. The latter sends projections to the hypothalamus and brainstem, allowing the expression of emotions through the modulation of autonomic and vegetative efferent motor systems. This was highlighted in this study by concomitantly monitoring the EDA, a marker of activation of the sympathetic system of the autonomic nervous system (Dawson et al.,
Scoring the intensity of the IAPS photographs by subjects is similar to that provided by the IAPS. The average scoring of intensity by our subjects is slightly lower to that of the IAPS norms (minus 0.5 points), which is similar to scoring carried out by our neighboring countries (Switzerland, Germany; Bradley and Lang,
Thanks to the IAPS norms, we were able to build a set of stimuli sensitive to intensity, devoid of bias relating to valence, and validated in our population of subjects.
In both sessions (inside and outside the MRI scanner), frequency of SCRs was greater for more intense stimuli (Group 3). Magnitude was only significantly higher for the more intense stimuli in the first session, in the MRI scanner. In contrast to data from other studies (Dawson et al.,
We obtained identical statistical results in both sessions on the relationship between stimuli intensity and amplitude of SCRs (no link during both sessions), and the relationship between stimuli intensity and frequency of SCRs (increased for Group 3 in the 2 sessions). The only difference in result between sessions is magnitude. The magnitude of SCRs is related to stimuli intensity only during the first phase, in the MRI scanner. There is a non-significant tendency during the second phase, outside the MRI scanner. One can draw several hypotheses to explain this difference. It is, first of all, a matter of calculation of magnitude, involving both the concept of response frequency and of amplitude. As amplitude was not sensitive to stimuli intensity in our study, it is likely that it is the frequency that makes variations in magnitude significant. Furthermore, it is also possible that the lack of significance of magnitude in the second session is related to a habituation effect to stimuli. Intra-subject stability of SCRs has been demonstrated 1 year later (Schell et al.,
The values of the magnitudes, amplitudes, and frequencies of SCRs that we recorded in the MRI scanner are lower than those recorded outside the MRI scanner (Figure
These results highlight the sensitivity of the amygdala to variations in the intensity of positive emotions, and in conjunction with results in the literature on negative emotions, show the role of the amygdala in the perception of emotional intensity.
However, the role of the amygdala has been identified in other areas, such as the behavior of consciousness, attentional mobilization (Pribram and McGuinness,
The study of emotions in healthy subjects will eventually lead to a better understanding of dysfunctions in the pathology of emotions. The amygdala has been studied in several psychiatric pathologies involving emotional disorders such as anxiety disorders and schizophrenia. However, as its roles are multiple, underpinned by its many connections with cortical and subcortical structures, the pathophysiology of the amygdala is not limited to emotional disorders themselves but in fact extends to areas such as memory, attention, decision-making, and cognition.
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.
This work was supported by Besancon University Hospital. We would also like to thank Holly Sandu and Nathalie Islam-Frenoy for providing linguistic help.
electrodermal system
International Affective Picture System
skin conductance response.