Browsing by Subject "causal models"

Aims: In earlier research it has been found that music can elicit strong emotional responses, and that mood affects the way they are processed. The brain basis of musical emotions has however been studied less than that of facial emotions. Also, there have been no studies on how depressed mood affects musical emotions in the brain. In the present study, both the effect of depressed mood state and that of depressive disorder on neural processing of musical emotions is studied. The aim is to identify brain regions affected, and to model the effective connectivity between these regions and the impact of depressed mood state and depressive disorder on this system. Methods: A functional magnetic resonance imaging (fMRI) experiment was conducted, in which 56 adult subjects listened to emotional (happy, sad, and fearful) music. The experiment consisted of two conditions: in the implicit condition the subjects were asked how many instruments were playing, and in the explicit which of the three emotions best characterised the musical excerpt. The subjects also completed the Montgomery-Åsberg Depression Rating Scale (MADRS) and the Profile of Mood States (POMS). The behavioural and imaging data were analysed both within the general linear model (GLM) to identify affected brain regions and the dynamic causal modelling (DCM) framework to model how sensory inputs enter the brain system and how experimental conditions modulate connections between specified brain regions. Results and conclusions: The subjects were mostly not clinically depressed (76%) and the MADRS scores were not correlated with neural activity in the brain. The POMS Depression scale was however associated with attenuated activity in the right posterior cingulate cortex (PCC) while listening to happy and fearful music under the implicit condition. As PCC has been associated with internally directed cognition and the management of brain's attention networks, this attenuation is likely to reflect the incongruence between stimuli and mood state, which would result in increased attention and/or a decrease in explorative cognitive activity. Comparisons of DCM models consisting of PCC, superior temporal gyrus (STG) and the amygdala indicated that auditory stimuli enter into this system via the auditory cortex in STG. The analyses could not determine whether the emotional content of the stimuli modulates connectivity between these regions, but MADRS and POMS scores were associated with amygdala connectivity. This is in line with the view that depressive disorder disrupts the amygdala's role in orienting to affective information.