ARTICLE IN BRIEF
Using a mathematical formula from signals derived from high-density electroencephalography recordings while patients are undergoing transcranial magnetic stimulation, investigators developed an index that they say helps distinguish between different levels of consciousness.
A team of scientists has developed a tool that can reliably differentiate levels of consciousness across a wide group of people — from healthy people during wakefulness, sleep, or under different sedatives; to patients in a locked-in state who can't move on command but who understand everything; to those who have been diagnosed as minimally conscious; to others in a permanent vegetative state.
If replicated and validated by others, the method — using a mathematical formula from signals derived from high-density electroencephalography (HD-EEG) recordings while patients are undergoing transcranial magnetic stimulation (TMS) — could be used in patients following traumatic brain injury or other conditions to assess their level of consciousness.
In the past decade, the field has tried different methods to tap consciousness but experts say this is the first time that scientists have offered such convincing evidence that they can separate out those who are conscious from those who are not — and do so through a formula that gives a single value from zero to one.
A NEW INDEX FOR BRAIN RESPONSE
Marcello Massimini, MD, PhD, a neurophysiologist at the University of Milan, and his colleagues at the University of Wisconsin and at the Coma Science Group in Liege have been conducting such studies for almost a decade. The recent work, published Aug. 14 in Science Translational Medicine, offers a new twist. They introduced a new index called the perturbational complexity index (PCI) that enables them to test how the brain responds to an environmental stimulus. This is a window into whether the brain picks up signals from the environment and integrates them into a meaningful experience. They use TMS to activate brain circuits and then calculate this response and compress the spatiotemporal pattern into a single value. Basically, they “zip” the information, as is commonly done in the case of a JPEG file. The less they can zip, the more information there is. This is how they arrive at a measure of complexity. One value — the PCI — can distinguish between consciousness and unconsciousness.
This PCI value, the scientists said, “reliably discriminated the level of consciousness” in single individuals.
In theory, Dr. Massimini explained, each conscious experience has specific features that differentiate it from other experiences and it is integrated into the singular experience. It is akin to different instruments in an orchestra that come together to create a unique piece of music.
The global response of the brain to the TMS perturbation relies on the ability of “multiple, functionally specialized areas of the thalamocortical system to interact rapidly and effectively to form an integrated whole,” the scientists explained in the Science paper.
“We faced a general problem,” Dr. Massimini said in an interview. “Since we don't know what really matters for consciousness in the brain, we normally rely on behavior. We assess consciousness by asking a question or by stimulating people and observing behavior. But what if people are disconnected from their environment and remain conscious, like patients with locked-in syndrome? Our goal was to find some aspects of brain activity and measure it independent of sensory processing.
“Neurophysiologically speaking, there is an incredible amount of interaction when every part of the brain is working as a whole,” he added.
The investigators have been working this model out for years. In the current experiment, they tested the reliability of the new algorithm — PCI — in a large database of 32 healthy people and 20 patients who had been in a coma and remain in various stages of consciousness — from a vegetative state to minimally conscious state (MCS) and locked-in. Healthy people were tested during wakefulness and sleep (rapid eye movement and non-rapid eye movement).
The PCI value was between zero and one. The closer a number is to zero the less likely the brain is processing the information from the TMS. “When the brain is able to integrate a lot of information, the PCI is high,” explained Dr. Massimini.
When they tested healthy patients on different sedatives they found that the PCI was lowest in the most powerful sedatives. REM sleep PCI values were higher than non-REM sleep.
They took 48 measurements from 20 brain-injured patients and compared them with the PCI values of the 32 healthy individuals. As predicted, PCI was lowest in patients in a vegetative state, falling within values of the healthy people in states of deep sleep and anesthesia. The two locked-in patients had PCI values that were on par with the healthy, awake subjects. The PCI values for MCS patients fell above a healthy person's score during sleep or anesthesia, that is, they had intermediate levels of consciousness.
The PCI values ranged from 0.44 to 0.67 in the healthy people during wakefulness and fell to 0.18 to 0.28 during non-REM sleep. During sedation, the scores ranged from a low of 0.21 to a high of 0.31. PCI values among the six patients in a vegetative state went from 0.19 to 0.31. MCS patients scored between 0.32 and 0.49 and the locked-in patients from 0.51 to 0.62.
Dr. Massimini said that their next step is to validate the measures in patients where there is ambiguity about the diagnosis. He said that such a tool can be used to diagnose and guide treatment. He said that they are also interested in figuring out whether there would be a way to determine complexity in the damaged brain. “Once we have a mechanism and know the critical changes in neurons that bring about consciousness and complexity, there may well be a way to increase complexity,” he said.
EXPERTS WEIGH IN
Scientists who study consciousness in brain-injured patients agree PCI is a measure that could someday be used clinically in intensive care units to establish levels of consciousness.
“It would be a major clinical advance if the EEG-TMS-PCI could provide prognostic information about brain function in the ICU,” said Nicholas D. Schiff, MD, the Jerold B. Katz professor of neurology and neuroscience at the Feil Family Brain and Mind Research Institute at Weill Cornell Medical College, who wrote an accompanying editorial in the same issue of Science Translational Medicine.
“What is important about the tool is that it gives you additional information that does not require intact sensory or motor channels. We do a lot of things to figure out awareness — fMRI, language, imagery, and observation, all of which require integration of the sensory system. But this is different. They bypass all of these systems and ring the brain like a bell” — that is, he explained, the patient is not called on to think or to respond to verbal commands; the brain's response to the TMS is what they are using to determine consciousness.
In an interview, Dr. Schiff added some important caveats that need to be worked out. “If a patient is on the boundary of MCS this tool won't help differentiate between conscious and unconscious states,” he said. Also, the score will fluctuate depending on when the patient is tested. “If a TBI patient is asleep the number would be the same as a person who is unconscious.”
He added that setting up the TMS-HD-EEG is not trivial. “You need a lot of trained people to do it right. But it is a great start.”
Anil Seth, PhD, professor of cognitive and computational neuroscience and co-director of the Sackler Centre for Consciousness Science at the University of Sussex in the United Kingdom, agrees. “If you apply TMS during conscious wakefulness, this pulse has a complicated spatial and temporal patterning. They quantify just how complex this response is. The approach is promising. It is specific and sensitive in determining levels of consciousness. If this stands up, it seems like an important step in identifying consciousness across a wide group of patients.”
“The findings suggest that normal consciousness occurs in the brain when long-range communication of information is possible. In states of decreased consciousness, altered brain function may prevent this communication,” said Hal Blumenfeld, MD, PhD, professor of neurology, neurobiology, and neurosurgery and director of the Yale Clinical Neuroscience Imaging Center at Yale University School of Medicine. “The investigators probe the normal human brain non-invasively using magnetic waves and then measure the brain's responses using a high density net of electrical sensors. Combining these powerful techniques can yield important insights into brain function.”
Benzi M. Kluger, MD, associate professor of neurology and psychiatry and director of the Movement Disorders Center and Director, Interdisciplinary Transcranial Magnetic Stimulation Laboratory at the University of Colorado Denver, said that TMS has been an exciting research tool but it wasn't clear whether it would have a clinical application. “This study tells us that we could use TMS to see things that you wouldn't see on a clinical exam. If this could be used to identify people who would eventually recover, that would be a big step forward for clinical neurology.”