Journal Articles
Abstract:
Functional brain imaging techniques provide a window into neural activity underpinning diverse cognitive processes, including visual perception, decision-making, and memory, among many others. By treating functional imaging data as a pattern-recognition problem, similar to face- or character-recognition, researchers have successfully identified patterns of brain activity that predict specific mental states; for example, the kind of an object being viewed. Moreover, these methods are capable of predicting mental states in the absence of external stimulation. For example, pattern-classifiers trained on brain responses to visual stimuli can successfully predict the contents of imagery during sleep. This research shows that internally mediated brain activity can be used to infer subjective mental states; however, it is not known whether more complex emotional mental states can be decoded from neuroimaging data in the absence of experimental manipulations. Here we show that brain-based models of specific emotions can detect individual differences in mood and emotional traits and are consistent with self-reports of emotional experience during intermittent periods of wakeful rest. These findings show that the brain dynamically fluctuates among multiple distinct emotional states at rest. More practically, the results suggest that brain-based models of emotion may help assess emotional status in clinical settings, particularly in individuals incapable of providing self-report of their own emotional experience.