Seeking the Neural Signature of Consciousness

In the 1971 film Johnny Got His Gun, the protagonist, Jon Bonham, lies helpless in a hospital bed after being hit by a mortar shell during the first World War. His physical injuries are horrendous—but more frightening to the viewer are his many attempts to communicate with others that he is, indeed, conscious, since he is unable to intentionally move or speak. He is trapped in his own body, helpless yet aware, and his physicians believe him to be in a vegetative state. Medicine has evolved considerably since World War I—and Bonham’s case is an extreme one—but it is still difficult for clinicians to effectively characterize the disorders of consciousness, including coma and persistent vegetative states, that can occur after severe brain injury. But now, researchers from the University of Cambridge may have identified a neural signature of consciousness in unresponsive patients at rest using electroencephalography (EEG) that may assist physicians with diagnosis and treatment.

Measuring consciousness

If a person has sustained a serious head injury and is comatose or unresponsive, clinicians attempt to determine the level of their consciousness using measures like the Coma Recovery Scale or the Glasgow Coma Scale. These type of scales allow doctors to systematically assess patients and look for signs of even minimal consciousness using various sensory modalities like vision, hearing, and movement, said  Srivas Chennu, a researcher at the University of Cambridge.

“These scales have been around for the last 10 to 15 years. In a sense, they are quite new,” he says. “But what the Coma Recovery Scale does is systematically breaks consciousness down into a patient’s awareness of the external world into different sensory modalities. And that helps physicians improve their identification of ‘hidden’ awareness in some patients and helped to identify patients who were thought to be vegetative but were actually minimally conscious.”

But despite the success of these scales, Chennu and colleagues wondered if some patients might be missed. “Even with the best scales, relying on behavior only may miss a significant proportion of minimally conscious patients. They would be in the minority, but it would still be a significant minority,” he says. “Some research suggests even 20% of patients, who are classified as vegetative on the coma recovery scale may have some sort of covert awareness. And we wondered if perhaps there was something going on in the brain networks of these patients that would help us better identify them.”

Applying graph theory

To test the idea, Chennu and colleagues used EEG, a non-invasive method of measuring brain activity by placing electrodes on the scalp, to measure brain activity at rest in 32 vegetative patients and 26 healthy adults. The researchers then used a technique of mathematical analysis called graph theory to look not only at the brain activity overall but also the strength of connections across different brain networks. It turned out that the connections mattered:  The researchers found highly abnormal network activity in most of the vegetative patients. But in three patients, they found network activity in the alpha band, or brain wave activity that represents network coordination and communication, that more closely matched the healthy adults. The results were published on October 16, 2014, in PLoS Computational Biology.

Anil Seth, co-director of the University of Sussex’s Sackler Centre for Consciousness Science, says the result fit well with the growing body of research on how the brain makes us aware. “There is an emerging perspective in our understanding of consciousness that it doesn’t have to do with activity in one particular area,” he says. “Rather, what’s necessary for consciousness is that different parts of the brain are contributing different aspects to some unified whole. That there’s an important interactivity among different regions, some kind of synchrony that leads to a conscious state.”

Chennu says this work has remarkable implications for our understanding of consciousness in general. “The differences we saw suggest there are potentially forms of consciousness that we’re still coming to terms with. We don’t know quite what sort of consciousness some of these patients have. But we do know that whatever is going on in their brains, despite the lack of behavioral response, is dramatically different from a truly vegetative patient. But is it like the consciousness that you and I take for granted?  We don’t know. We are still trying to figure out how to assay the subject of consciousness. And this kind of analysis may help us better do that.”

Diagnostic or prognostic?

Tristan Bekinschtein, a colleague of Chennu’s at the University of Cambridge and co-author of the study, says that this work may do more than just help us better understand the neural signature of consciousness—it may also help us improve treatment for people who are minimally conscious. And it can be done in an inexpensive and non-invasive way.

“This data gives you a different way to think about what the patient can do. You can do the classic tests to see if the patient is responding or not. But even if the patient shows nothing, you may see a well-informed information processing network using the EEG,” he says. “And if that is the case, regardless of any other test, the clinician can say, ‘This guy has the potential to process information. I don’t know why he’s not processing information, but maybe we can find out out.’  And hopefully, knowing that capacity is there will make physicians keep their eyes open and less likely to write the patient off.”

Chennu and Bekinschtein are also looking at whether alpha band activity may be prognostic, helping to predict which patients are more likely to recover from disorders of consciousness over time.

“Studies have shown that brain activity measured by positron emission tomography (PET) in the regions that support these alpha band oscillations were more likely to recover later,” says Chennu. “So there is some hint that the presence of these working networks can prognosticate recovery, but that finding remains to be tested in EEG physiology. We’d like to replicate that, and we plan to follow these patients to see if we can replicate that. We’re just not there yet.”