BioMed Research International

BioMed Research International / 2014 / Article
Special Issue

Translational Neuroimaging of the Mood and Anxiety Disorders

View this Special Issue

Research Article | Open Access

Volume 2014 |Article ID 906104 | 9 pages |

Abnormal Early Gamma Responses to Emotional Faces Differentiate Unipolar from Bipolar Disorder Patients

Academic Editor: Yong He
Received15 Oct 2013
Revised30 Jan 2014
Accepted01 Feb 2014
Published13 Mar 2014


This study investigates the cortical abnormalities of early emotion perception in patients with major depressive disorder (MDD) and bipolar disorder (BD) using gamma oscillations. Twenty-three MDD patients, twenty-five BD patients, and twenty-four normal controls were enrolled and their event-related magnetoencephalographic responses were recorded during implicit emotional tasks. Our results demonstrated abnormal gamma activity within 100 ms in the emotion-related regions (amygdala, orbitofrontal (OFC) cortex, anterior insula (AI), and superior temporal pole) in the MDD patients, suggesting that these patients may have dysfunctions or negativity biases in perceptual binding of emotional features at very early stage. Decreased left superior medial frontal cortex (smFC) responses to happy faces in the MDD patients were correlated with their serious level of depression symptoms, indicating that decreased smFC activity perhaps underlies irregular positive emotion processing in depressed patients. In the BD patients, we showed abnormal activation in visual regions (inferior/middle occipital and middle temporal cortices) which responded to emotional faces within 100 ms, supporting that the BD patients may hyperactively respond to emotional features in perceptual binding. The discriminant function of gamma activation in the left smFC, right medial OFC, right AI/inferior OFC, and the right precentral cortex accurately classified 89.6% of patients as unipolar/bipolar disorders.

1. Introduction

Mood disorders, including major depressive disorder (MDD) and bipolar disorder (BD), are among the most debilitating diseases worldwide and with a lifetime prevalence estimated about 20.8% [1]. Guidelines from the Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV) [2], characterize the manic/hypomanic episodes as an index for distinguishing bipolar from unipolar disorders, but misdiagnosis of bipolar as unipolar could occur particularly when the BD patients come to the hospital in depressive episodes [3, 4]. Identification of disorder biomarkers disclosed from neuroimage researches could improve diagnostic accuracy and clinical treatment outcomes of bipolar disorder [4]. Precious neuroimaging studies have proposed that dysfunction of facial expression perception is one of core impairments in the MDD and BD patients [58]. Hence quantitative measurements of neural responses to emotional stimulation may facilitate accurate diagnosis and better treatment outcomes of the MDD and BD patients.

Depressed patients have an attentional bias toward negative stimuli (easier to attract patients’ attention) [9], which are more inclined to interpret neutral faces as sad [10, 11]. Stimulation with high arousal levels would activate the negative motivational withdrawal system more strongly than the positive approach system [12], which is called the negativity bias [13]. It implicates that withdrawal from negative stimuli is more critical to survival than approach of pleasant or neutral stimuli [13, 14]. To achieve this goal, humans “early” allocate attentional resources to negatively valenced stimuli in an efficient way, as proposed in the automatic vigilance model of emotion [15], which leads to delayed attentional disengagement. Hence we speculate that the depression symptoms in mood disorders may associate with their increased negativity bias, which results in dysfunction of early rapid processing of resources allocation.

Implicit emotional paradigms, by which participants attend to judge nonemotional perceptual features (e.g., gender) rather than emotional cues, have been considered an effective way to explore neural substrates of facial emotional processing [16, 17]. Perceptual processing of emotion-laden stimuli engages several critical brain regions, including the amygdala, prefrontal cortex, orbitofrontal cortex (OFC), anterior cingulate cortex (ACC), and anterior insula (AI) [18]. Previous studies showed that the abnormal activity of these regions in patients with affective disorder may be related to their specific symptoms, including anhedonia (easy to be unhappy and sad) and emotional instability [5]. However, there is limited understanding of dysfunction of neurobiological basis underlying “early” emotion perception. The present study aimed at elucidating whether the MDD and BD patients have impaired or biased perception of emotional facial features at very early stage of emotional processing.

The current study measured gamma activity by magnetoencephalography (MEG) and adopted implicit emotional paradigms to investigate early facial emotion perception in BD and MDD patients, compared with healthy controls and to distinguish these two affective disorders by discriminant analyses. Gamma-band responses have been implicated to be a mechanism of perceptual binding and strongly synchronized about 100 ms after sensory processing, reflecting integration processing of perceived features at very early stage [8, 19, 20]. Müller et al. [21] suggests that spatial distribution of gamma oscillations, including limbic, temporal, and frontal cortices may be linked to neural substrates of binding emotional information. Gamma oscillations can provide a potential index to explore regional brain abnormalities of early emotional processing in affective disorders. Previous studies have showed that subcortical and cortical regions activate within 100 ms by high temporal resolution technique, such as event-related potentials (ERPs) and magnetoencephalography (MEG) [22, 23]. Our previous study also demonstrated alterations of gamma activity (50–150 ms) during early emotion processing in the MDD and BD [7]. We in this study further tested that alterations of cortical gamma responses to early perception of emotional facial expression were distinct between the MDD and BD patients, which can be a potential index to differentiate patients with unipolar from bipolar disorders.

2. Materials and Methods

2.1. Subjects

Seventy-two participants were recruited from the Department of Psychiatry at Taipei Veterans General Hospital in this study, including twenty-three MDD patients (mean age , nine males), twenty-five BD patients (mean age , ten males), and twenty-four age- and gender-matched healthy controls (NC, mean age , nine males). The three groups did not show significant differences in age (a one-way ANOVA, ) and gender (2 × 3 contingency table analysis, ). All subjects were right-handed as assessed by the Edinburgh Handedness Inventory. The diagnosis of MDD and BD was confirmed by a structured interview based on the Diagnostic and Statistical Manual for Mental Disorders (DSM-IV) criteria (American Psychiatric Association, 1994). Before MEG data acquisition, psychiatric and mood symptoms of all patients were assessed with the Young Mania Rating Scale (YMRS) and the 17-item Hamilton Rating Scale for Depression (HAMD). For details, see Table 1. The NC subjects underwent the Mini International Neuropsychiatric Interview before participation in the study to confirm the absence of past or current psychiatric symptoms. Each participant signed informed consent forms approved by the Institutional Review Board at Taipei Veterans General Hospital.


Gender (male : female) 9 : 14 11 : 149 : 15
Age (years)35.96 (9.58)36.80 (11.38)36.62 (11.36)
Age of onset (years)26.65 (9.26)28.80 (9.71)
Duration of illness (years)9.30 (7.64) 7.92 (6.18)
Number of manic episodes 3.08 (2.34)
Number of depressive episodes 4.26 (3.61) 4.12 (2.17)
Number of major depressive episodes 2.61 (1.08) 3.00 (1.00)
Young Mania Rating Scale (YMRS) 0.77 (1.60) 1.44 (2.02)
Hamilton Depression Rating Scale (HAMD) 9.09 (8.22) 6.80 (5.29)

Except for the gender variable, all other variables are presented as mean (SD).

2.2. Stimuli and Experimental Design

Face images with duration of 1.5 sec (72 trials for each emotion and four emotions in total, including neutral, sad, happy, and angry faces) were displayed randomly. Visual stimuli were exhibited at the center of a back-projected translucent screen, which was located 100 cm in front of the subject, and subtended 14° (width) by 17° (height) of visual angle. Subjects were instructed to perform a gender discrimination task by lifting their left/right index finger for male/female face images, respectively, while a response cue was displayed. All subjects practiced the test before their MEG signals were recorded.

2.3. MEG and MRI Recordings

Event-related MEG data were recorded by a whole-head 306-channel neuromagnetometer (Vectorview 306, Elekta Neuromag, Finland) with a sampling rate of 1000 Hz and a 0.03~330 Hz bandpass filter. Trials containing deflections exceeding 9000 fT/cm or contaminated by eye movements were excluded for the source analysis. The signal space projection method [24] was applied to remove urban and device interference in the recorded MEG data. The T1-weighted MRI (magnetic resonance images) of each individual was acquired by a 1.5 T GE Signa Excite scanner using an 8-channel phased-array head coil with 3D fast spoiled gradient recalled echo (3D FSPGR, TR = 8.67 ms, TE = 1.86 ms, inversion time = 400 ms, matrix size = 256 × 256 × 124, and voxel size = 1.02 × 1.02 × 1.5 mm3). To facilitate precise coregistration of the MEG data and structural MRI, three anatomical landmarks (nasion and left and right preauriculars) were localized with Isotrak 3D digitizer (Polhemus Navigation Sciences, Colchester, Vermont, USA).

2.4. MEG Source Analysis

For each emotion, the noise-free MEG data were filtered at a frequency band of 35 to 55 Hz (gamma). These gamma-band signals were then analyzed through a beamforming method [25] to estimate cortical activity index of emotional evoked responses, which was denoted as the gamma-band activation index (GBAI). The GBAI map with an isotropic voxel size of 4 mm in the whole-brain was obtained by estimating the ratio between estimated signals of the active state (a 30-ms window) and those of the control state (from 300 ms to 200 ms before stimulus onset) for each voxel. This study focused on investigating the difference of brain responses between patients with MDD and BD in emotion processing during the first 100 milliseconds after stimulus onset (from 30 to 120 ms with 5 ms shift).

For group analysis, the individual T1-weighted MRIs were first transformed into the Montreal Neurological Institute (MNI) space (a standard stereotactic space) with an isotropic spatial resolution of 2 mm by the BIRT software [26]. The obtained deformation field was then applied to transform the individual GBAI maps obtained above into the MNI space for further group analysis. A one-way ANOVA (F(2, 69) = 7.65, uncorrected , cluster size = 100) and two two-sample t-tests (t(47) = 3.87 for BD versus NC group, t(45) = 3.88 for MDD versus NC group, uncorrected 7) were conducted at each time point in a voxel-wise manner by using the statistical parametric mapping software (SPM2, The intersection areas between the survived voxels obtained from the ANOVA and t-test analyses were extracted and the mean of GBAI within each area with cluster size ≥30 was calculated for the following correlation and discriminant analyses.

2.5. Correlation and Discriminant Analyses

Pearson correlation coefficients were used to assess the relationship between the abnormal regional GBAIs and symptomatic/demographic data in patient groups. The correlation was determined to be significant at least 10 ms (three continuous maps, e.g., 30, 35, and 40 ms).

Abnormal regional GBAIs of the MDD and BD patients obtained from the above-mentioned source analysis procedure were extracted as possible features for differentiating these two patient groups. A two-stage discriminant analyses were performed to identify distinguishable feature variables (abnormal regional GBAIs) and their weightings by evaluating their contributions in distinguishing patient groups (MDD versus BD). In the first stage, a stepwise linear discriminant procedure was used to select the predictors of the model, which can be best able to distinguish between these two patient groups. At each step, one variable was considered at a time and Wilks’ Lambda values of the variables in the model were used to determine this variable to be a predictor or not. The threshold of Wilks’ Lambda was set at 0.25 for setting retention in the model, which was based on previous Monte Carlo simulation studies [27]. This procedure iteratively repeated for each variable until there was no further improvement in discriminability of the model.

The second stage was a canonical discriminant analysis which was used to determine the weightings of those variables selected from the first step. The resolved standardized weight of each variable (regional GBAI) reflected its relative discriminating efficiency. The accuracy of the derived discriminant function from the two-stage discriminant analyses was assessed by leave-one-out cross validation for the whole patient cohort. Furthermore, a two-sample t-test was used to test whether the mean values of the discriminant function for the patient groups were different. Finally, a Fisher Exact Test was performed to evaluate the statistical significance of the classification accuracy [28].

3. Results

Overall, both patient groups displayed regional gamma hyperactivity compared to the NC group, but only the MDD patients exhibited the diminished gamma activity, as listed in Table 2. Overall, we found diminished gamma responses at very early time points (30–70 ms after stimulus onset) and elevated gamma responses at later time points (80–115 ms). Figures 1 and 2 showed the brain regions with abnormal gamma activity at peak time points (shown in Table 2) in the MDD and BD patients, respectively.

Group difference Emotion Brain regionBATime (ms)Coordinate (mm) Cluster size
DurationPeakx, y, z at peak

MDD < NCSadR anterior insula/Inferior orbitofrontal cortex BA4740–706026, 14, −204.94457
R superior temporal pole/parahippocampal cortexBA3855–706030, 16, −264.7339
R amygdala55–655522, −4, −124.4730
HappyL superior medial frontal cortex BA830–4035−6, 42, 564.664
R medial orbitofrontal cortex BA10/11/3230–403516, 52, −124.46366
MDD > NCSadR precentral cortex BA495–11510522, −24, 624.87217
BD < NCSad/happy/angry/neutral       
HappyR middle temporal cortex BA2290–1009046, −56, 204.471
AngryR middle/inferior occipital cortexBA1880–1009532, −96, −45.15258

The significant voxels were performed by a one-way ANOVA of three groups (uncorrected ) and between-group comparisons with a Bonferroni adjustment (uncorrected ), all values >3.87. The cluster size denotes the number of voxels and coordinates are in MNI space. BA: Brodmann area; R: right; L: left.

The MDD patient showed decreased gamma responses to sad and happy faces within 70 ms and increased gamma responses to sad faces around 100 ms, relative to the NC group (Table 2). No significant difference of brain responses to angry or neutral faces was found. The decreased gamma responses to sad faces were in the right anterior insula/inferior OFC (40–70 ms), the right superior temporal pole/parahippocampus (55–70 ms), and the right amygdala (55–65 ms). The hypoactivity responding to happy faces was located in the right superior/medial OFC and left superior medial frontal cortex during 30–40 ms. On the other hand, the right precentral/postcentral cortex (95–115 ms) of the MDD patients was more activated to sad faces, compared to the NCs.

As to the BD patients, only increased responses were found in comparison with the NC group, including the right middle temporal cortex (90–100 ms) to happy faces and the right middle/inferior occipital cortex (80–100 ms) to angry faces. (Table 2). No significant difference between the BD and NC groups in response to sad and neutral facial expressions was found.

3.1. Correlation between Abnormal Regional GBAI and Symptomatic Data

The correlations between the abnormal regional GBAIs and the symptomatic and demographic data of the patient groups were assessed. Only the HAMD score of the MDD patients was found to be negatively correlated with the GBAI of the left superior medial frontal cortex (r = −0.441, at 30 ms; r < −0.521, at 35 and 40 ms), as shown in Figure 3. In the BD patients, no significant correlation between symptomatic/demographic data and abnormal regional GBAIs was found.

3.2. Discriminant Analysis

The stepwise linear discriminant analysis identified four brain regions with the most distinguishing capability between the MDD and BD patients (F4,43 = 9.77, ), including the left superior medial frontal cortex (happy 35 ms), the right medial orbitofrontal cortex (happy 35 ms), the right anterior insula/inferior OFC (sad 60 ms), and the right precentral cortex (sad 105 ms). The outputs of the discriminant functions between these two patient groups were significantly different (t47 = 20.87, ). Figure 4 illustrated the distribution of the discriminant function scores in the MDD (mean = 0.38, se = 0.32) and BD patients (mean = −3.43, se = 0.38). The results of leave-one-out cross validation showed that no MDD patient was misclassified and only five out of twenty-five BD patients (80%) were misclassified into the MDD group. Overall, 43 of 48 patients were correctly categorized with a prediction accuracy of 89.6% (Fisher’s Exact Test, P 3.28 × 10−9).

4. Discussion

Our results demonstrated the distinct patterns of gamma oscillatory abnormalities in the MDD and BD patients in responses to emotional facial expression during early perceptual processing. The MDD displayed more deficits in the frontal and limbic regions, including amygdala, OFC, and anterior insula, than the BD. Among these regions, the patterns of gamma activation in the left smFC, right mOFC, right AI/inferior OFC, and the right precentral cortex can accurately classify 89.6% of patients into their diagnosed categories.

Notably, in our study there were two BD patients who were initially diagnosed as the MDD patients and then confirmed as the bipolar disorder two weeks later. Our data from the discriminant analyses showed that these two BD patients were correctly classified using their brain signals measured at the first week (the circle marks in Figure 4) although the patients in this study were not drug-naive or drug-free, which could have a confounding effect on the brain signals. These results suggest that the gamma responses to emotional faces can provide a useful objective index to differentiate the BD patients from the MDD patients and may become a potential biomarker to assist in diagnosis.

Our data showed diminished gamma activity at the amygdala, OFC, and insula in response to sad faces during early emotion perception in MDD patients. This finding may unravel three possible neural mechanisms underlying emotion perception in human brain. First, gamma oscillation could be an emotion-evoked oscillation. Being considered as a mechanism of feature binding [11, 12], early sensory-evoked gamma oscillations were reliably found in various modalities [29]. Early visual evoked gamma-band response within 100 ms after stimulation is proposed to be sensitive to attentional and perceptual factors [30]. The finding of the aberrant gamma responses to emotional faces within 100 ms in the present study indicates that human brain integrates emotional facial features in a very effective and rapid way at the early stage of perceptive processing revealed by gamma oscillations.

Second, the frontoinsular cortex as well as amygdala has been reported to be involved in fast processing of salient information obtained from emotional facial features [31]. The amygdala is proposed to be engaged very early (within 100 ms) in processing negative faces as disclosed by neurophysiological studies using MEG [22] and intracranial electroencephalography [23]. Adolphs [32] reported that humans process emotional facial expressions in perception with simple and highly salient features within 120 ms including the amygdala and OFC. Our finding of diminished gamma activity within 100 ms in the amygdala, anterior insula, and OFC in the MDD patients may indicate their impairments or inefficiency in rapid processing or integrating salient emotion.

Finally, the MDD patients were reported to have a negativity bias (easier to attract patients’ attention toward negative stimuli) [9], especially toward sad stimuli. We speculated that the negative-affect bias of depressed patients may be associated with the OFC activity in response to sad facial expressions. The OFC, a key region of the top-down facilitation model, is suggested to rapidly extract low spatial frequency components of inputs (at around 50 ms) from visual or subcortical cortices through the magnocellular route to generate possible candidates of objects [33, 34]. Eliminated OFC activity may result in disrupted top-down information processing. Our results of the diminished activity of the OFC within 50 ms in the MDD patients may suggest a neural evidence of top-down modulation of negative-affect bias in the MDD patients while facing a sad expression.

Our data also indicated that the more severe depression symptoms the MDD patients had, the more eliminated the left smFC responses to happy faces were. Our results were in line with the previous studies [6, 35], which also showed the deactivation of the left prefrontal cortex in depressed patients. Mitterschiffthaler et al. [36] also indicated that reduced medial frontal responses to positive valence stimuli in depressed patients were related to abnormalities of positive emotion processing. Our finding suggests that decreased activity in the smFC underlies irregular positive emotion processing in patients with major depression, which may be one of neural substrate candidates related to anhedonia in depression.

In the BD patients, our finding showed elevated activation in visual regions responded to emotional facial expressions at around 100 ms after visual stimulation. In line with our previous findings of the enhanced occipitotemporal gamma oscillations in the BD patients in sensor-space analysis [8], in this study we found the abnormal occipital or temporal regions of the BD patients when they perceived only the happy and angry faces but not the sad and neutral faces. Happy faces consist of more changeable facial features (e.g., mouth) than negative and neutral faces [37]. Extracting the negative valence from angry expressions is easier than that from sad expressions [38, 39]. Happy and angry faces are high arousal stimuli relative to sad and neutral faces [40, 41]. Hence, the findings of the hyperactivation in the occipital and temporal regions of the BD patients, only to the happy and angry faces but not to the sad and neutral faces, reflect that dysfunction of perceptual processing in the BD patients may be associated with detection of changeable as well as high-arousal facial features. The finding of the hyperactivity of visual regions around 100 ms in the BD patients indicates that the BD patients have altered visual perception of emotional features which may lead to dysfunction of the subsequent cognitive functions.

Other oscillations (theta, alpha, and beta) play different roles for neuronal functions. A review paper [42] summarizes that theta oscillations are a key mechanism of memory processing in the hippocampus (the main brain area related to theta waves); alpha oscillations are related to function of inhibition in the motor cortex; beta oscillations are associated with functions of motor control and attention in cortical structures. Previous studies indicated that high gamma oscillations (>60 Hz) are related to cognitive and perceptual processes [43, 44], but its significance or role remains unclear [45, 46]. On the other hand, low gamma oscillations (30–50 Hz) have been well documented as a crucial mechanism of perceptual binding and object/face representation in numerous human studies [29, 4749]. The “binding problem” addresses the physiological mechanisms responsible for combining different features in a visual scene to form a coherent percept [29]. The present study focused on investigating the low gamma responses to emotional faces within 100 ms to reveal the dysfunction of perceptual binding in emotional feature processing of the MDD and BD patients.

5. Conclusions

Our results demonstrated that abnormal activation within 100 ms of the MDD patients in the emotion-related regions (amygdala, inferior/medial OFC, AI, and superior temporal pole) responded to emotional facial expressions, which suggests that the MDD patients may have dysfunctions or negativity biases in perceptual binding of salient emotional features at very early stage. In the BD patients, our finding showed that abnormal activation in visual regions (inferior/middle occipital and middle temporal cortices) responded to emotional facial expressions very early within 100 ms, which supports that the BD patients may hyperactively or sensitively respond to emotional features in perceptual binding. Decreased responses to happy faces in the MDD patients at the left smFC were correlated with their serious depression symptoms, which may support that decreased activity in the smFC underlies irregular positive emotion processing in patients with major depression. The discriminant function of four variables, including gamma activation in the left smFC, right mOFC, right AI/inferior OFC, and the right precentral cortex, accurately classified 89.6% of patients as unipolar/bipolar disorders. These findings indicate different impairments of brain regions in the MDD and BD patients during early facial emotional perception and this abnormal regional gamma activity can be a potential index to distinguish these two mood disorders.

Conflict of Interests

The authors declare that there is no conflict of interests regarding the publication of this paper.


This study was supported in part by National Science Council (NSC97-2752-B-010-003-PAE and NSC100-2629-B-010-001) and Taipei Veterans General Hospital (V97ER1-001) and Brain Research Center, National Yang-Ming University. The authors thank Chih-Che Chou, Chou-Ming Cheng, and Ying-Chia Lin for their assistance with data collection.


  1. R. Kessler, P. Berglund, O. Demler, R. Jin, K. R. Merikangas, and E. E. Walters, “Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the national comorbidity survey replication,” Archives of General Psychiatry, vol. 62, no. 6, pp. 593–602, 2005. View at: Publisher Site | Google Scholar
  2. A. P. Association and DSM-IV. APATFo, Diagnostic and Statistical Manual of Mental Disorders: DSM-IV, Amer Psychiatric, Arlington, Va, USA, 1994.
  3. D. Muzina, D. Kemp, and R. McIntyre, “Differentiating bipolar disorders from major depressive disorders: treatment implications,” Annals of Clinical Psychiatry, vol. 19, no. 4, pp. 305–312, 2007. View at: Publisher Site | Google Scholar
  4. M. L. Phillips and E. Vieta, “Identifying functional neuroimaging biomarkers of bipolar disorder: toward DSM-V,” Schizophrenia Bulletin, vol. 33, no. 4, pp. 893–904, 2007. View at: Publisher Site | Google Scholar
  5. M. L. Phillips, W. C. Drevets, S. L. Rauch, and R. Lane, “Neurobiology of emotion perception II: implications for major psychiatric disorders,” Biological Psychiatry, vol. 54, no. 5, pp. 515–528, 2003. View at: Publisher Site | Google Scholar
  6. N. S. Lawrence, A. M. Williams, S. Surguladze et al., “Subcortical and ventral prefrontal cortical neural responses to facial expressions distinguish patients with bipolar disorder and major depression,” Biological Psychiatry, vol. 55, no. 6, pp. 578–587, 2004. View at: Publisher Site | Google Scholar
  7. J. R. C. D. Almeida, A. Versace, A. Mechelli et al., “Abnormal amygdala-prefrontal effective connectivity to happy faces differentiates bipolar from major depression,” Biological Psychiatry, vol. 66, no. 5, pp. 451–459, 2009. View at: Publisher Site | Google Scholar
  8. T. Y. Liu, J. C. Hsieh, Y. S. Chen, P. C. Tu, T. P. Su, and L. Chen, “Different patterns of abnormal gamma oscillatory activity in unipolar and bipolar disorder patients during an implicit emotion task,” Neuropsychologia, vol. 50, no. 7, pp. 1514–1520, 2012. View at: Publisher Site | Google Scholar
  9. I. H. Gotlib, E. Krasnoperova, D. N. Yue, and J. Joormann, “Attentional biases for negative interpersonal stimuli in clinical depression,” Journal of Abnormal Psychology, vol. 113, no. 1, pp. 127–135, 2004. View at: Publisher Site | Google Scholar
  10. C. Bourke, K. Douglas, and R. Porter, “Processing of facial emotion expression in major depression: a review,” Australian and New Zealand Journal of Psychiatry, vol. 44, no. 8, pp. 681–696, 2010. View at: Publisher Site | Google Scholar
  11. A. Stuhrmann, T. Suslow, and U. Dannlowski, “Facial emotion processing in major depression: a systematic review of neuroimaging findings,” Biology of Mood & Anxiety Disorders, vol. 1, pp. 1–17, 2011. View at: Publisher Site | Google Scholar
  12. S. E. Taylor, “Asymmetrical effects of positive and negative events: the mobilization-minimization hypothesis,” Psychological Bulletin, vol. 110, no. 1, pp. 67–85, 1991. View at: Google Scholar
  13. J. T. Cacioppo, W. L. Gardner, and G. G. Berntson, “The affect system has parallel and integrative processing components: form follows function,” Journal of Personality and Social Psychology, vol. 76, no. 5, pp. 839–855, 1999. View at: Publisher Site | Google Scholar
  14. S. Kousta, D. P. Vinson, and G. Vigliocco, “Emotion words, regardless of polarity, have a processing advantage over neutral words,” Cognition, vol. 112, no. 3, pp. 473–481, 2009. View at: Publisher Site | Google Scholar
  15. F. Pratto and O. P. John, “Automatic vigilance: the attention-grabbing power of negative social information,” Journal of Personality and Social Psychology, vol. 61, no. 3, pp. 380–391, 1991. View at: Google Scholar
  16. H. Critchley, E. Daly, M. Phillips et al., “Explicit and implicit neural mechanisms for processing of social information from facial expressions: a functional magnetic resonance imaging study,” Human Brain Mapping, vol. 9, no. 2, pp. 93–105, 2000. View at: Google Scholar
  17. J. Scheuerecker, T. Frodl, N. Koutsouleris et al., “Cerebral differences in explicit and implicit emotional processing-an fMRI study,” Neuropsychobiology, vol. 56, no. 1, pp. 32–39, 2007. View at: Publisher Site | Google Scholar
  18. L. Pessoa, “On the relationship between emotion and cognition,” Nature Reviews Neuroscience, vol. 9, no. 2, pp. 148–158, 2008. View at: Publisher Site | Google Scholar
  19. C. Tallon-Baudry, O. Bertrand, F. Peronnet, and J. Pernier, “Induced γ-band activity during the delay of a visual short-term memory task in humans,” Journal of Neuroscience, vol. 18, no. 11, pp. 4244–4254, 1998. View at: Google Scholar
  20. E. Rodriguez, N. George, J. Lachaux, J. Martinerie, B. Renault, and F. J. Varela, “Perception's shadow: long-distance synchronization of human brain activity,” Nature, vol. 397, no. 6718, pp. 430–433, 1999. View at: Publisher Site | Google Scholar
  21. M. M. Müller, A. Keil, T. Gruber, and T. Elbert, “Processing of affective pictures modulates right-hemispheric gamma band EEG activity,” Clinical Neurophysiology, vol. 110, no. 11, pp. 1913–1920, 1999. View at: Publisher Site | Google Scholar
  22. Q. Luo, T. Holroyd, M. Jones, T. Hendler, and J. Blair, “Neural dynamics for facial threat processing as revealed by gamma band synchronization using MEG,” NeuroImage, vol. 34, no. 2, pp. 839–847, 2007. View at: Publisher Site | Google Scholar
  23. W. Sato, T. Kochiyama, S. Uono et al., “Rapid amygdala gamma oscillations in response to fearful facial expressions,” Neuropsychologia, vol. 49, no. 4, pp. 612–617, 2011. View at: Publisher Site | Google Scholar
  24. M. A. Uusitalo and R. J. Ilmoniemi, “Signal-space projection method for separating MEG or EEG into components,” Medical and Biological Engineering and Computing, vol. 35, no. 2, pp. 135–140, 1997. View at: Google Scholar
  25. Y. S. Chen, C. Y. Cheng, J. C. Hsieh, and L. F. Chen, “Maximum contrast beamformer for electromagnetic mapping of brain activity,” IEEE Transactions on Biomedical Engineering, vol. 53, no. 9, pp. 1765–1774, 2006. View at: Publisher Site | Google Scholar
  26. J.-X. Liu, Y.-S. Chen, and L.-F. Chen, “Fast and accurate registration techniques for affine and nonrigid alignment of MR brain images,” Annals of Biomedical Engineering, vol. 38, no. 1, pp. 138–157, 2010. View at: Publisher Site | Google Scholar
  27. M. C. Costanza and A. Afifi, “Comparison of stopping rules in forward stepwise discriminant analysis,” Journal of the American Statistical Association, vol. 74, no. 368, pp. 777–785, 1979. View at: Publisher Site | Google Scholar
  28. R. M. Chapman, J. W. McCrary, M. N. Gardner et al., “Brain ERP components predict which individuals progress to Alzheimer's disease and which do not,” Neurobiology of Aging, vol. 32, no. 10, pp. 1742–1755, 2011. View at: Publisher Site | Google Scholar
  29. C. Tallon-Baudry and O. Bertrand, “Oscillatory gamma activity in humans and its role in object representation,” Trends in Cognitive Sciences, vol. 3, no. 4, pp. 151–162, 1999. View at: Publisher Site | Google Scholar
  30. C. S. Herrmann, M. H. J. Munk, and A. K. Engel, “Cognitive functions of gamma-band activity: memory match and utilization,” Trends in Cognitive Sciences, vol. 8, no. 8, pp. 347–355, 2004. View at: Publisher Site | Google Scholar
  31. J. Fan, X. Gu, X. Liu et al., “Involvement of the anterior cingulate and frontoinsular cortices in rapid processing of salient facial emotional information,” NeuroImage, vol. 54, no. 3, pp. 2539–2546, 2011. View at: Publisher Site | Google Scholar
  32. R. Adolphs, “Neural systems for recognizing emotion,” Current Opinion in Neurobiology, vol. 12, no. 2, pp. 169–177, 2002. View at: Publisher Site | Google Scholar
  33. M. Bar, “A cortical mechanism for triggering top-down facilitation in visual object recognition,” Journal of Cognitive Neuroscience, vol. 15, no. 4, pp. 600–609, 2003. View at: Publisher Site | Google Scholar
  34. M. Bar and M. Neta, “The proactive brain: using rudimentary information to make predictive judgments,” Journal of Consumer Behaviour, vol. 7, no. 4-5, pp. 319–330, 2008. View at: Publisher Site | Google Scholar
  35. J. B. Henriques and R. J. Davidson, “Left frontal hypoactivation in depression,” Journal of Abnormal Psychology, vol. 100, no. 4, pp. 535–545, 1991. View at: Publisher Site | Google Scholar
  36. M. T. Mitterschiffthaler, V. Kumari, G. S. Malhi et al., “Neural response to pleasant stimuli in anhedonia: an fMRI study,” NeuroReport, vol. 14, no. 2, pp. 177–182, 2003. View at: Publisher Site | Google Scholar
  37. J. M. Leppänen and J. K. Hietanen, “Positive facial expressions are recognized faster than negative facial expressions, but why?” Psychological Research, vol. 69, no. 1-2, pp. 22–29, 2004. View at: Publisher Site | Google Scholar
  38. S. Murphy, “The nonconscious discrimination of emotion: evidence for a theoretical distinction between affect and emotion,” Polish Psychological Bulletin, vol. 32, no. 1, pp. 9–15, 2001. View at: Google Scholar
  39. P. Winkielman, K. C. Berridge, and J. L. Wilbarger, “Unconscious affective reactions to masked happy versus angry faces influence consumption behavior and judgments of value,” Personality and Social Psychology Bulletin, vol. 31, no. 1, pp. 121–135, 2005. View at: Publisher Site | Google Scholar
  40. J. A. Russell and A. Mehrabian, “Evidence for a three-factor theory of emotions,” Journal of Research in Personality, vol. 11, no. 3, pp. 273–294, 1977. View at: Google Scholar
  41. S. Droit-Volet, S. Brunot, and P. M. Niedenthal, “Perception of the duration of emotional events,” Cognition and Emotion, vol. 18, no. 6, pp. 849–858, 2004. View at: Publisher Site | Google Scholar
  42. P. J. Uhlhaas, C. Haenschel, D. Nikolić, and W. Singer, “The role of oscillations and synchrony in cortical networks and their putative relevance for the pathophysiology of schizophrenia,” Schizophrenia Bulletin, vol. 34, no. 5, pp. 927–943, 2008. View at: Publisher Site | Google Scholar
  43. C. Grützner, P. J. Uhlhaas, E. Genc, A. Kohler, W. Singer, and M. Wibral, “Neuroelectromagnetic correlates of perceptual closure processes,” Journal of Neuroscience, vol. 30, no. 24, pp. 8342–8352, 2010. View at: Publisher Site | Google Scholar
  44. J.-P. Lachaux, N. George, C. Tallon-Baudry et al., “The many faces of the gamma band response to complex visual stimuli,” NeuroImage, vol. 25, no. 2, pp. 491–501, 2005. View at: Publisher Site | Google Scholar
  45. S. Ray and J. H. R. Maunsell, “Different origins of gamma rhythm and high-gamma activity in macaque visual cortex,” PLoS Biology, vol. 9, no. 4, Article ID e1000610, 2011. View at: Publisher Site | Google Scholar
  46. P. J. Uhlhaas, G. Pipa, S. Neuenschwander, M. Wibral, and W. Singer, “A new look at gamma? High- (>60 Hz) γ-band activity in cortical networks: function, mechanisms and impairment,” Progress in Biophysics and Molecular Biology, vol. 105, no. 1-2, pp. 14–28, 2011. View at: Publisher Site | Google Scholar
  47. C. M. Gray, P. Konig, A. K. Engel, and W. Singer, “Oscillatory responses in cat visual cortex exhibit inter-columnar synchronization which reflects global stimulus properties,” Nature, vol. 338, no. 6213, pp. 334–337, 1989. View at: Google Scholar
  48. W. Lutzenberger, F. Pulvermüller, T. Elbert, and N. Birbaumer, “Visual stimulation alters local 40-Hz responses in humans: an EEG-study,” Neuroscience Letters, vol. 183, no. 1-2, pp. 39–42, 1995. View at: Publisher Site | Google Scholar
  49. W. Singer and C. M. Gray, “Visual feature integration and the temporal correlation hypothesis,” Annual Review of Neuroscience, vol. 18, pp. 555–586, 1995. View at: Google Scholar

Copyright © 2014 T. Y. Liu et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

1349 Views | 496 Downloads | 7 Citations
 PDF  Download Citation  Citation
 Download other formatsMore
 Order printed copiesOrder

We are committed to sharing findings related to COVID-19 as quickly and safely as possible. Any author submitting a COVID-19 paper should notify us at to ensure their research is fast-tracked and made available on a preprint server as soon as possible. We will be providing unlimited waivers of publication charges for accepted articles related to COVID-19. Sign up here as a reviewer to help fast-track new submissions.