Researchers map path to consciousness after brain injury

A new AI tool can spot signs of consciousness in brain-injured patients four-to-eight days earlier than standard bedside tests.

The system, called SeeMe, tracks microscopic facial movements that are invisible to the naked eye to detect awareness in patients labelled as “unresponsive”.

Up to 25 per cent of patients diagnosed as unresponsive may in fact be conscious but unable to show it – a state known as cognitive motor dissociation (CMD).

Researchers at Stony Brook University’s Renaissance School of Medicine developed SeeMe to address this gap in neurology and critical care.

Dr Sima Mofakham is senior author, associate professor and vice chair of research in neurosurgery, and assistant professor in electrical and computer engineering.

Mofakham said: “We developed SeeMe to fill the gap between what patients can do and what clinicians can observe,” said

“Just because someone can’t move their limbs or speak doesn’t mean they aren’t conscious. Our tool uncovers those hidden physical efforts by patients to show they are conscious.”

The technology uses high-resolution video and computer vision to measure involuntary facial reactions to verbal prompts such as “open your eyes” or “show me a smile”.

Machine learning then analyses the tiny movements that clinicians usually cannot detect.

In a clinical study of 37 patients with acute brain injury and coma, SeeMe detected purposeful movement up to four days before doctors recognised any visible response.

Patients with early SeeMe-detected responses were more likely to regain consciousness and achieve better functional outcomes at discharge.

The results build on the 2024 Bodien et al. study, which found that up to 25 per cent of patients labelled unresponsive in intensive care may retain brain function that standard tests fail to detect.

Misdiagnosis can delay rehabilitation for patients who might otherwise recover.

Dr Chuck Mikell, neurosurgeon, co-lead investigator and clinical associate professor, said: “This is not just a new diagnostic tool, it’s a potential prognostic marker.

“Families often ask us how long it will take for a loved one to wake up, or if they ever will.

“This study helps us answer those questions with more confidence, grounded in data, not just experience or instinct.

“We can use this information to personalise care, guide families, and optimise rehabilitation efforts.”

The ethical stakes are high. Misdiagnosis can lead to inappropriate withdrawal of care, reduced access to rehabilitation, and missed treatment opportunities.

SeeMe is described as non-invasive, inexpensive and scalable, requiring only a camera and open-source software – making it potentially viable in hospitals with limited resources.

Dr Mofakham said: “This kind of work shows the future of medicine lies at the intersection of disciplines.

“We aim to turn complex data into tools that help doctors make faster, better decisions when every hour counts.”

The team is planning larger trials and seeking regulatory approval.

They hope eventually to integrate SeeMe with EEG, which records brain activity, and other data to create a multi-modal platform for monitoring consciousness.