A recent study led by researchers Sima Mofakham and Chuck Mikell, MD, from the Renaissance School of Medicine at Stony Brook University, has introduced an artificial intelligence tool that could change how clinicians detect consciousness in patients with traumatic brain injury (TBI). The findings were published in Nature Communications Medicine.
The new AI system, named SeeMe, was developed to identify covert signs of awareness in patients who appear unresponsive. According to the research team, up to a quarter of brain-injured individuals labeled as “unresponsive” may actually be conscious but unable to physically demonstrate it—a condition known as cognitive motor dissociation (CMD).
“We developed SeeMe to fill the gap between what patients can do and what clinicians can observe,” said Mofakham, senior author of the study and associate professor at Stony Brook University. “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.”
In a clinical trial involving 37 acute brain injury and coma patients, SeeMe used high-resolution video analysis and computer vision technology to track subtle facial movements made in response to verbal instructions such as “open your eyes” or “show me a smile.” These reactions are typically too slight for medical staff to notice unaided.
The researchers found that SeeMe detected purposeful movement four days earlier on average than standard clinical assessments. In some cases, detection came as much as eight days sooner.
“This kind of work shows the future of medicine lies at the intersection of disciplines, as we begin to see more applications of AI and engineering in medicine. With such an approach, we aim to turn complex data into tools that can help doctors make faster and better decisions for patients when every hour counts,” said Mofakham.
Patients whose responses were identified early by SeeMe were more likely to regain consciousness before discharge and had improved functional outcomes compared with those who did not show early signs according to the tool.
“This is not just a new diagnostic tool, it’s a potential prognostic marker,” said Mikell. He added: “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 personalize care, guide families, and optimize rehabilitation efforts.”
The authors noted that misdiagnosis of unresponsive states may result in premature withdrawal of care or missed opportunities for therapy. They also highlighted ethical considerations related to patient treatment decisions.
Referencing a 2024 New England Journal of Medicine study by Bodien et al., which reported that 15–25 percent of ICU patients diagnosed as unresponsive might retain higher-level brain function undetectable by current bedside exams, the Stony Brook team emphasized the need for objective assessment tools like SeeMe.
According to Mofakham and Mikell, SeeMe is noninvasive and affordable; it requires only basic camera equipment and open-source software—features that could make it accessible even in hospitals with limited resources.
Looking ahead, larger trials are planned along with efforts toward regulatory approval. The researchers envision integrating SeeMe into regular intensive care monitoring alongside other methods such as EEGs.
Funding for this project came from institutional seed grants supporting collaboration between Stony Brook University’s Departments of Neurosurgery and Electrical and Computer Engineering.