Innovative Tool Reveals Hidden Signs of Consciousness in Brain Injury Patients

Researchers at Stony Brook University have developed and tested a new computer vision technology called SeeMe, capable of detecting subtle, voluntary facial movements in patients with acute brain injuries who are unresponsive or in a coma. This innovative tool identified low-amplitude movements—such as eye opening, tongue protrusion, and smiling—in response to auditory commands days before traditional clinical assessments could recognize overt signs of consciousness.

The study, published in tions Medicine, involved 37 comatose patients and 16 healthy volunteers aged 18 to 85. The patients had Glasgow Coma Scale scores of 8 or lower, with no prior neurological conditions. During testing, SeeMe analyzed facial pores at a resolution of approximately 0.2 mm, tracking movement vectors as subjects listened to commands like "open your eyes," "stick out your tongue," and "show me a smile." These responses were recorded on video, then evaluated against blinded reviews and standard clinical examinations.

Results demonstrated that SeeMe could detect signs of consciousness an average of 4 days earlier than conventional clinical assessments—specifically, eye-opening responses appeared 9.1 days post-injury compared to 13.2 days as observed in clinical exams. The tool successfully identified eye responses in over 85% of patients and mouth movements in nearly 94% of cases where the mouth was unobstructed, often preceding overt signs detected by clinicians.

Further, the technology showed promising potential in predicting patient outcomes, with the intensity and frequency of responses correlating to recovery scores. A deep neural network trained on the collected data achieved 81% accuracy in identifying specific command responses, notably eye movements.

This breakthrough suggests that many patients considered unresponsive may exhibit covert consciousness through subtle facial cues, which went unnoticed with standard bedside checks. Early detection of such signs can influence clinical decision-making, guide rehabilitation efforts, and facilitate future brain-computer interfaces for communication.

The study underscores the importance of integrating advanced AI and computer vision techniques into neurodiagnostics, offering a non-invasive, widely accessible method to better understand the hidden cognitive states of brain injury patients. Source: [https://medicalxpress.com/news/2025-09-seeme-hidden-consciousness-brain-injury.html]