A bedside brain-computer interface has allowed clinically unresponsive patients to follow commands using only their thoughts. The breakthrough could fundamentally reshape how intensive care teams assess and communicate with people suffering from severe acute brain injuries.
The system builds on years of work led by Jan Claassen, medical director of the neurocritical care unit at NewYork-Presbyterian/Columbia University Irving Medical Center. His team previously used standard EEG technology paired with machine learning to uncover hidden consciousness in 15% of unresponsive patients, a finding that stunned the critical care community.
Those patients could not move or speak, yet their brain waves revealed attempts to comply with instructions like “make a fist” or “wiggle your toes.” Researchers later determined the disconnect stems from a breakdown between brain regions governing cognition and those controlling movement. Patients with this condition, called cognitive motor dissociation, stand a much better chance of regaining consciousness and recovering daily function.
Now Claassen’s team has gone further. They attached EEG leads to 41 unresponsive patients within days of hospitalization. A beeping tone played through headphones while a recorded voice asked each patient to change the beep’s frequency by imagining opening and closing their hands. Electrical signals from the brain fed into a bedside computer that analyzed them in real time. When patients adjusted the tone successfully, the voice offered encouragement.
Roughly one in three participants gained control of the beeps using thought alone. That included every patient already identified as having cognitive motor dissociation through separate testing. Remarkably, this occurred within days of catastrophic brain injury.
Previous research demonstrated fMRI could extract yes or no answers from one unresponsive patient, but fMRI remains impractical for routine ICU use. EEG, meanwhile, already monitors brain-injured patients daily. The only novel component sits in the software that decodes signals and modulates feedback.
Families reported the technology helped them feel connected to their loved one. They suggested adding personalized voices, culturally meaningful content, and alerts for pain, hunger, or awareness. Claassen aims to refine the system so patients can one day tell doctors directly what they feel.
