Researchers at Stony Brook University have developed SeeMe, an AI-powered computer vision tool that detects subtle facial movements in coma patients in response to voice commands, revealing signs of covert consciousness that are often invisible to standard clinical exams.

This AI algorithm can identify tiny, purposeful facial gestures—such as micro-smiles or slight eye openings—days before larger, observable responses appear, enabling earlier detection of consciousness.

In a study involving 37 recent brain injury patients, SeeMe detected eye-opening movements on average four days earlier than clinicians, demonstrating its potential to provide faster, more objective assessments.

Patients exhibiting more frequent and larger facial movements tend to have better recovery prospects, highlighting the importance of early detection for prognosis and treatment planning.

Early identification of covert consciousness through SeeMe could significantly influence medical decisions, guiding rehabilitation efforts and reassuring families about patient prospects.

This technology offers a faster, more accessible alternative to traditional neuroimaging methods, which are often time-consuming and require specialized skills.

The ability to detect signs of awareness early could help clinicians tailor treatments, potentially leading to improved outcomes for patients with brain injuries.

Future research aims to explore whether patients can use facial movements to answer yes-or-no questions, opening new possibilities for communication with non-verbal patients.

SeeMe can identify low-amplitude, purposeful gestures, such as micro-smiles or slight eye openings, that are typically invisible to the naked eye but are clinically significant.

This breakthrough paves the way for future innovations like brain-computer interfaces, which could enable communication with patients previously considered entirely unconscious.

The phenomenon of 'covert consciousness' describes patients who are awake and aware but unable to communicate or show outward signs of consciousness, making detection crucial for prognosis.

Unlike traditional neuroimaging, which is time-consuming and requires specialized skills, SeeMe offers a rapid, accessible method for detecting consciousness through video analysis.

The study demonstrated that SeeMe detected eye-opening responses up to 4.1 days earlier than clinicians and identified responses in more patients overall, marking a significant advancement in early diagnosis.