CMU Study Identifies Emotions from Brain Activity

futurelab default header
One of the ongoing controversies in neuromarketing is how well current techniques can identify specific emotions. While there’s general agreement that attention and emotional engagement can be tracked, identifying specific emotions with confidence has been elusive. Now, researchers at Carnegie Mellon University have published a new study showing the ability to identify emotions with an accuracy “well above chance” using fMRI.


Although hundreds of papers have been published linking brain activity to emotional states, this is the first study to offer results showing the ability to identify the specific emotional state of the subject based on brain activity. The key difference, according to the paper, is that the new technique goes beyond identifying specific areas of the brain and linking them to a specific emotion. Rather, the CMU scientists identified patterns of activity across many areas of the brain that represented particular emotional states.
One interesting aspect of the research was the way the researchers gauged the emotions the subjects were experiencing. Rather than the usual “self-report” approach of asking subjects how they feel, the experimenters used “method” actors from CMU’s drama school to create the emotions.
This research bodes well for the future of neuromarketing, even if the current techniques aren’t ready for immediate commercial application. The lead author of the study, Karim Kassam, sees the potential, noting, “It could be used to assess an individual’s emotional response to almost any kind of stimulus, for example, a flag, a brand name or a political candidate.”
Kassam talks about the work in more detail in this video.