247 SeeMe: A Brain-Computer Interface for Characterizing the Recovery of Consciousness

INTRODUCTION: Recovery of consciousness after traumatic brain injury (TBI) is uncertain and hard to detect. We developed SeeMe, a real-time brain-computer interface to measure low-amplitude voluntary movements that are hard to detect with the naked eye. SeeMe uses computer vision to analyze facial microexpressions responding to auditory commands such as “Open your eyes.” For the first time, SeeMe integrates neural activity, represented in scalp electroencephalography (EEG) and/or electrocorticography (ECoG), with a behavioral response time series in one device. METHODS: This study is registered under the ClinicalTrials.gov ID number NCT06083441 and approved by the Stony Brook University Institutional Review Board. Informed consent was obtained from each patient’s legally authorized representative. We sought to enroll patients at Stony Brook Hospital who suffered acute brain injuries (TBI, cerebral hemorrhage) and subsequently became comatose. While on Natus Xltek EEG, SeeMe played four commands in 10 trial blocks: a control command (“Today is a sunny day”) and three action commands (“Stick out your tongue,” “Open your eyes,” “Show me a smile”). SeeMe integrated EEG/ECoG and behavioral time series and identified biomarkers of responsiveness. RESULTS: Six patients were enrolled in this study (83% male, mean age 47.8 (18.7) years old), with eleven recording sessions total. Five patients demonstrated EEG beta power modulation after commands, with one showing beta burst signaling. This beta modulation correlated with the detection of intentional responses by SeeMe and was absent in the control command. Four of the five patients with beta power modulation went on to follow commands before discharge. CONCLUSIONS: Our results illustrate the relationship between the beta power modulation and command following and showcase SeeMe as a brain-computer interface for characterizing brain-behavior relationships.