Neuromagnetic Responses to Facial Expression Images in Patients with Major Depressive Disorder

• Main Authors: Hsiao-Ping Shen, 沈曉萍
• Other Authors: Li-Fen Chen
• Format: Others
• Language: en_US
• Published: 2010
• Online Access: http://ndltd.ncl.edu.tw/handle/51278952975978215895

碩士 === 國立陽明大學 === 生物醫學資訊研究所 === 98 === Background: To differentiate the perception and identification processes between different affective categories can aid the understanding of neuropathology in major depressive disorder (MDD). The present study proposed a novel technique by utilizing independent component analysis and cortical source imaging to investigate the different impaired neural substrates of emotional responses to negative and positive facial images in MDD. Methods: Nineteen MDD patients and nineteen age- and gender-matched normal controls (NC) were recruited and performed a gender discrimination task with facial expression images. The event-related magnetoencephalographic signals were recorded and processed through independent component analysis, beamformer-based source imaging analysis, and clustering. Four time components were specified for further statistical analysis Results: We found that the MDD patients had significantly increased cortical activation of inferior frontal and occipital regions evoked by negative emotion, compared with positive emotion, at all time components. The MDD patients showed under-activation of posterior in both emotions than the NC group, especially cingulated cortex and anterior frontal pole in negative emotion and hippocampus and lingual region in positive emotion, respectively. Only at the third component (NC: 140ms, MDD: 125ms), the MDD patients showed over-activation of temporal cortex than NC group in negative emotion. Conclusions: Our results demonstrated the significantly different neuronal circuitries between MDD and NC groups corresponding to different facial expression stimuli. These findings suggest the dysfunction of identification and generation of emotional responses to emotional salient information in the MDD patients. The proposed methods could obtain the significant neuronal substrates of emotion processing which act as seeds to further examine the model of sub-networks.