Detecting Confusion Using Facial Electromyography

摘要

Objective: The ability to detect confusion using facial electromyography (EMG) was evaluated.rnBackground: Facial EMG has been useful in detecting various emotional states. Another state potentially detectable from facial musculature is confusion. Confusion reflects a loss of understanding or, in a dynamic environment, a loss of situation understanding or situation awareness.rnMethod: For this study, 24 participants listened to a confusing and a neutral audio passage while they were monitored with EMG. Electrical activity of the right and left corrugator supercilii, the right zygomaticus major, and right depressor anguli oris was recorded. Participants also reported whether they felt confused during the experimental passage. Finally, they were asked to pose facial expressions. Two raters assessed responses in the graphed EMG data for evidence of activation near a confusing sentence in the passage.rnResults: EMG revealed confusion in 21 of 24 (87.5%) participants, even in some participants who did not admit confusion. Posed expressions of confusion differed somewhat from spontaneous expressions elicited during the experimental passage.rnConclusion: EMG was effective in detecting confusion, regardless of operator self-report. Not all muscles in the face were important for the detection of confusion: The zygomaticus major led to a large number of false positives.rnApplication: EMG can be an effective addition to a sensor suite designed to monitor the cognitive state of operators of systems.