If people with brain injuries no longer react recognizably to speech, this is considered a bad sign. Those affected are therefore regularly asked to move their hand or stick out their tongue. If they repeatedly do not respond, it is difficult to predict whether they will recover: usually the injuries are so severe that they do not wake up again. But the EEG can give reason for hope, as a study in "The Lancet Neurology" shows. If there are signs of awareness in the EEG, the chances of recovery are better.
In the summer of 2021, a research group from New York examined almost 200 patients with acute brain injuries who no longer showed a visible reaction to a response. The doctors used electroencephalography to record the brain activity before and after the speech and compared the brain waves with machine learning. In 14 percent of them, they observed a so-called cognitive-motor dissociation: while the affected people appeared unconscious on the outside, their brain activity changed when they were spoken to.
A good sign, as it turned out: Most of these patients were accessible again after three months. After a year, 41 percent were so much better that they could do large parts of the day alone; In contrast, only 10 percent of those who had not registered EEG did not register any evidence of consciousness.
So far, there has been no reliable method to predict which patients will recover, says study leader Jan Claassen, a neurologist at Columbia University. "In the future, the dissociation between cognitive and motor functions could be another factor in the prognosis." The hidden consciousness reveals more about the further course than other established characteristics such as age or the cause of the brain damage. Currently, however, only a few centers with EEG would be looking for covert awareness. The group therefore wants to develop an AI software that can infer consciousness from the brain waves. You have already published the algorithm for free use.
As early as 2019, Claassen reported on patients with colleagues in the "New England Journal of Medicine", in whom brain activity suggested a certain degree of consciousness. Even then, the forecast in these cases was significantly better. However, the sample was too small for reliable conclusions.
