AI Detects Hidden Signs of Consciousness in Coma Patients Days Before Doctors

AI Detects Hidden Signs of Consciousness in Comatose Patients Earlier than Doctors

Patients with traumatic brain injuries can experience “covert consciousness”—a state where they are aware but unable to move or communicate. A recent study published in Communications Medicine reveals that artificial intelligence can identify subtle facial movements indicating consciousness days before clinicians notice them.

Background on Covert Consciousness

Covert consciousness was first identified in 2006 when an unresponsive patient showed brain activity similar to healthy volunteers during mental imagery tasks. More recently, neuroimaging studies found that about 25% of behaviorally unresponsive patients actually possess covert consciousness. However, these imaging techniques are complex, time-consuming, and not routinely used in clinical settings. Instead, doctors rely on visual exams—checking for eye-opening, command following, or reflexive responses—to assess consciousness.

Using AI to Track Facial Movements

The research team, led by computational neuroscientist Sima Mofakham at Stony Brook University, developed an AI tool named SeeMe. They recorded videos of 37 comatose patients and analyzed their facial movements at a granular level, down to individual pores. Patients were given simple commands like “open your eyes” or “stick out your tongue.” SeeMe detected subtle, targeted facial movements that were invisible to the naked eye.

• Eye-opening responses were identified in 30 out of 36 patients.
• Mouth movements were detected in 16 out of 17 patients with suitable videos.
• In several cases, these small movements appeared 4 to 8 days before clinicians observed any signs of consciousness.

These findings suggest that consciousness returns gradually, with small facial cues preceding larger, more obvious movements.

Clinical Implications

Detecting consciousness earlier has practical benefits. Patients showing larger and more frequent facial movements tended to have better recovery outcomes, indicating that this AI-based method could assist in prognosis. Early detection informs treatment decisions—whether to pursue aggressive therapies or palliative care—and helps families and medical teams make more informed choices during critical periods.

Additionally, starting rehabilitation sooner is linked to improved motor function, and earlier recognition of consciousness could accelerate access to these therapies.

Future Directions

Consciousness recovery after brain injury is often unpredictable and fluctuates over time. While this study tracked patients up to six months post-discharge, there is potential for AI tools like SeeMe to detect awareness in long-term care patients currently classified as unconscious.

The next research phase aims to determine whether patients can use specific facial movements to answer yes-or-no questions. This could enable communication with patients who otherwise cannot express themselves, opening new ethical and care possibilities.

For professionals interested in AI applications in healthcare, exploring machine learning’s role in neurological assessment can provide valuable insights. More on AI training and courses can be found at Complete AI Training.