In March of this year, the beloved singer-songwriter Joni Mitchell was rushed to the hospital after being found unconscious on the floor in her Bel Air home. Subsequently, long-time friend Leslie Morris filed a petition to become Mitchell's legal guardian, stating that the singer "remains unconscious and unable to make any responses, and is therefore unable to provide for any of her personal needs."

Statements issued on Mitchell's official website deny that Mitchell is in a coma, claiming instead that "she comprehends, she's alert, and she has her full senses. A full recovery is expected." But Mitchell's doctor, the neurosurgeon Paul Vestra, has confirmed that her mental faculties are severely impaired and that she will unlikely be unable to attend the court hearing over guardianship, which is scheduled in July.

The conflicting information about Mitchell's health status reveals the stigma attached to coma and other conditions in which consciousness is impaired. In such situations, complex ethical questions regarding end-of-life decisions inevitably arise, ones that most societies are still struggling to answer.

"Clinically, consciousness refers to two quite distinct things - wakefulness and awareness," explains neuroscientist Adrian Owen of the University of Western Ontario in a recent public lecture. "But the problem for people working on consciousness is that none of us can agree on what it actually means."

Owen is less interested in defining consciousness than he is in measuring it. He's also keen on finding ways to detect hidden signs of consciousness and determining if patients have conscious awareness, exactly what they are aware of, and how much of their mental faculties they retain. Owen's is one of several research groups that are pioneering new ways of doing this, and although the work is still in its infancy, it is already helping clinicians to diagnose consciousness disorders more accurately.

Awake, Asleep, or Neither?

Our understanding of consciousness has been hampered by our historically rudimentary ways of understanding its absence. We can easily determine if someone is awake, because their eyes will probably be open. Wakefulness can also be detected if one's eyes are closed, because the awake and sleeping brain are each characterised by distinct patterns of brainwaves, which can be recorded with electroencephalography (EEG), using electrodes placed onto the scalp.

Awareness is much harder to detect, because the only way we have of knowing whether or not someone is aware is by asking them. Comatose patients, for example, have their eyes closed at all times, and are neither awake nor aware. Patients in the vegetative state, however, often show signs of wakefulness. "They often open their eyes and look around the room and they have intact sleep-wake cycles," Owen says, "but they won't fixate on anything in particular and, importantly, never respond to a command, so they're often referred to as awake but not aware."

This poses a major problem for clinicians working with patients who have consciousness disorders because, while minimally conscious patients are more likely to recover, they do not show outward signs of awareness. The problems with diagnosis thus make it extremely difficult to identify those whose conditions might improve.

Owen and his colleagues pioneered the use of functional neuroimaging to detect signs of consciousness in patients diagnosed as being in a persistent vegetative state, and to communicate with them. The method was first described in a landmark 2006 paper, and involves scanning patients' brains while asking them a series of questions that require a "yes or no" answer. Upon entering the scanner, the patients are instructed to imagine playing a game of tennis if they want to answer "yes" or to imagine walking around their apartment if their answer is "no."

Each task produces different brain activation patterns that be detected by the scanner - one activates the supplementary motor cortex, which is involved in planning movements, and the other activates the hippocampus and surrounding areas, which play critical roles in spatial navigation.

Remarkably, Owen and his colleagues found that a significant proportion of apparently vegetative patients can actually follow these commands and respond to the questions by imagining one scenario or the other. The ability to do so indicates that certain of their mental abilities - especially attention, memory, and language comprehension - are intact, and that they are, in fact, aware of what is going on around them.

"All of these things are part and parcel of what we call conscious awareness, things that we all use day every day just to get through our activities," Owen says. "We've seen a lot of patients, all of them diagnosed as being in the vegetative state, and about one in five of them are entirely unresponsive at the bedside but will reliably produce these activity patterns in the scanner."

On the basis of these results, Owen and his colleagues conclude that around 20% of patients with these disorders are misdiagnosed as being in a persistent vegetative state when they are actually minimally conscious. For many, this is a matter of life or death: Court cases that decide whether hydration and nutrition should be withdrawn from apparently vegetative patients are becoming increasingly common around the world, and such decisions often hinge on the question of whether or not the patient is conscious.

Brain scanning is expensive, and moving patients to have these tests can be hugely problematic, so Owen and his colleagues have developed a cheaper, portable EEG-based version of the test. They want to test as many patients as they can, and to that end have built what they call the EEJeep, which they use to visit patients in their homes. And they believe that some patients may be unable to follow commands despite retaining some awareness, so have developed a "passive" test based on the shared experience and brain activity people have when watching the same movie together.

Defining Consciousness

Despite these important advances, this research tells us very little about what consciousness actually is, or about how the brain generates it. Consciousness is at once familiar to us all, and deeply mysterious. We are all familiar with the content of our consciousness, and how those contents change as time passes, and we all lose consciousness every night when we go to sleep, only to regain it in the morning. Yet, neuroscientists and philosophers alike are still struggling to find an adequate definition of consciousness.

When brain scanning technologies first emerged in the 1990s, some researchers began using them to identify the neural correlates of consciousness - activity patterns in specific parts of the brain that are associated with certain aspects of awareness - but this did not explain the relationship between brain activity and experience or why we only become aware of some aspects of the brain's workings but not others.

Neuroscientists are increasingly viewing the brain as a complex network of inter-connected modules, and so there is growing interest in mapping the long-range neural pathways linking them. They have also come to believe that the synchronized activity of large groups of brain cells - which produces brain waves - is important for information processing and that the synchronization or de-coupling of brain wave frequencies between inter-connected brain regions is likely important for the flow of information between them.

But many say that we still have no real understanding of consciousness. "I often hear that we haven't made much progress and still don't know what we're talking about," says Anil Seth, director of the Sackler Centre for Consciousness Studies at the University of Sussex, "but I think that's simply wrong, because we've seen huge developments in understanding."

One such development is integrated information theory (IIT), developed by neuroscientist Giulio Tononi of the University of Wisconsin-Madison, which explains consciousness in terms of the amount of information that is shared amongst brain regions. IIT states that consciousness is graded, and equates the amount of information being integrated with the level of conscious experience, such that conscious awareness increases with the amount of integration taking place. It predicts that any sufficiently complex system that integrates information could have "bits of consciousness."

Thus, most animals are conscious to a greater or lesser extent, and the internet, with its billions of interconnected computers worldwide, could conceivably be conscious in some sense. Yet a supercomputer simulation of the human brain, which is only partly integrated, could not.

"This is an interesting theory that allows us to make all sorts predictions," Seth says. For example, it predicts that general anesthetics make us lose consciousness by reducing information integration in the brain to below a certain critical level and that integration is reduced or otherwise disrupted in consciousness disorders. Both of these predictions turn out to be accurate.

Marcello Massimini of the University of Milan and his colleagues developed a test based on Tononi's theory. It involves using transcranial magnetic stimulation to interfere with the brain's electrical activity and then measuring the complexity of the brain's response. Massimini and his colleagues have performed the test on healthy participants while they are awake, asleep, and at different levels of sedation by anesthetic, as well as in comatose, vegetative, and minimally conscious patients. They've shown that it can reliably determine the level of consciousness in each of these states.

Similarly, researchers in France have devised a way to measure how much information is being shared between pairs of scalp electrodes during EEG sessions. Using it to record brainwaves both in patients with consciousness disorders and in healthy controls, they found that unconscious patients' brains do indeed exhibit lower levels of what they called "global information sharing" than those of the others.

These methods could potentially be developed into cheap bedside tests that allow for more accurate diagnosis of consciousness disorders. Broader access to portable EEG tests promises to revolutionize how these patients are diagnosed and cared for.

"We've made a great deal of progress and discovered a lot, but it's a very hard problem, and it's not going to just go away so easily," Seth says. As new ways of measuring consciousness are developed further, we will become better equipped to deal with its disorders. And although we may still be a long way from understanding consciousness, by studying its various forms, we now seem to have a loose grasp of it.