The patient opens her eyes, but they are unfocused. She is awake yet apparently unaware of anything going on in the hospital room around her. After the accident, she lies in her bed, unresponsive, day after day. What is she thinking?

Soon we may be able to communicate with such "locked-in" minds--trapped in bodies that no longer respond to their mental control. In a blitz of publicity last fall, a team of British researchers announced they had imaged the brain of one of their "vegetative" patients and discovered that she was in fact conscious and aware. Now that same team has developed a way to ask yes-or-no questions of such patients. The idea is radical: we might soon be able to reach a number of people, including 250,000 Americans, who suffer from consciousness disorders--patients who, until now, had been considered beyond treatment.

"We are now able to detect when somebody is consciously aware, when existing clinical methods have been unable to provide that information," says Adrian Owen of the University of Cambridge, leader of the team of researchers who imaged the woman's brain as she responded to doctors' requests that she imagine such activities as playing tennis. Because of recent advances in imaging technology, patients "can literally communicate without having to say or do anything," Owen says.

A Shift in Thinking
"People have felt until now that this patient group isn't worth investing in. The attitude has been, 'There's nothing that can be done,'" Owen adds. Decades ago the medical community provided nothing more than palliative care for patients with disorders of consciousness who could not wake up or who were not aware of their surroundings. These brain-damaged patients were kept clean and comfortable until they died.

Once in a while, however, one of these patients would recover unexpectedly. When someone woke up out of a decade-long coma, the revival would be considered a miracle or, at the very least, a medical mystery. There seemed to be no way to determine if a patient with brain damage would come to or not. The only thing to do was wait and see.

But beginning in the 1970s, the scientific field of neurorehabilitation came into existence. Rehab centers, where patients could receive treatment from specialists, sprang up around the country. Doctors began to consider each brain injury individually, tailoring unique drug regimens and physical therapies in an attempt to improve each patient's condition. That was the best they could do, however: traditional magnetic resonance imaging (MRI)--used since the 1980s to map the structure of the brain or other areas inside the body--made it possible for doctors to see physical damage to the brain but did not allow them to examine its activity.

Then, in the 1990s, with the advent of functional MRI (fMRI) scans, it became possible to study activity in living brains. Functional MRI allows researchers to see which areas of the brain are most active during thought processes, which is how Owen and his colleagues determined that their vegetative patient was indeed aware and responding to their commands. Slowly, neuroscientists' understanding of brain damage began to move forward. Brain-damaged patients were no longer automatically considered lost causes but rather victims of a condition for which there might someday be a cure.

"Functional imaging is really the first imaging technique that has allowed us to look at the inner cognitive workings of patients who have disorders of consciousness," says Joy Hirsch, a neuroscientist at Columbia University. In 1992 scientists discovered they could use an MRI scanner to map changes in blood flow to different areas of the brain, signaling which parts of the brain were working during any given thought process or sensory stimulation. In the subsequent decade, researchers determined the difference in the fMRI patterns of willful thought and passive response to stimuli, a crucial distinction when examining the brain of a patient whose state of consciousness is unknown. Now fMRI technology has improved such that researchers can give patients commands and analyze their responses within a minute rather than a month. The result: we are on the verge of communicating with patients who only a few years ago would have been considered brain-dead.

Of course, not all patients can improve: some simply do not have enough brain structure left. "We've seen several recent cases that tell us that in some of these patients there is some chance of recovery," Owen explains. "But certainly not in all patients." This was the case with Terri Schiavo, a permanently vegetative Florida woman who became the center of a political debate in 2005 when her parents challenged her husband's decision to remove her feeding tube and let her die. A computed tomographic (CT) scan showed that much of her brain had atrophied, and doctors were unanimous in their opinion that she would not recover.

Determining a brain-damaged patient's prognosis is not always so cut and dried. The brain is a fragile organ; it can be damaged in many different ways, most of which are poorly understood by science. Whereas a number of patients might regain partial or complete use of their faculties, others truly are permanently injured with no hope of recovery. In a few cases, the victims might be aware of their surroundings but unable to respond. Still others are unconscious and unaware. The difficulty lies in determining which patients are which.

The first step is getting a general understanding of the patient's state of mind. Clinicians divide disorders of consciousness into three categories: coma, in which a patient is neither awake nor responsive; vegetative, in which a patient is awake but unresponsive; and minimally conscious, in which a patient is awake and responds to stimuli but has limited capacity to take willful actions. Typically doctors make these categorizations by observing a patient at the bedside. By this method alone, a patient thought to be vegetative could actually be aware.

"It's really a conundrum. The way that consciousness is typically measured is by basically asking somebody to tell you that they are conscious," Owen says. "So if someone wasn't unconscious but couldn't respond and tell you that, they would be classed as unconscious."

In Owen's team's case study, reported in the September 8, 2006, issue of the journal Science, the researchers asked the vegetative patient to imagine herself doing various tasks, including walking through the rooms of her home, while they scanned her brain using fMRI. The resulting images [see box on opposite page] showed that her response matched that of healthy test subjects--she understood the commands and intentionally decided to comply.

But analyzing the massive volume of data generated by an fMRI scan takes time. When fMRI was first developed, it took up to several months to interpret one scan. As recently as early 2006, when Owen's team scanned the patient's brain, data analysis took many days. "That eureka moment didn't come as she was lying in the scanner," Owen states. "Two weeks later we realized she had indeed been playing tennis in her head."

Now, Owen reports, fMRI technology has advanced to the point where researchers can interpret the data from a scan in 30 or 40 seconds. This breakthrough opens up the possibility of "reading" a person's thoughts at a given moment, enabling a locked-in patient to "speak" with only his or her mind.

New Therapies
Owen and his colleagues hope that one day the new fMRI techniques they are developing will assist doctors in determining which patients are aware but trapped in an unresponsive body, thereby providing a more reliable indication of patients' potential for recovery. The researchers report that their patient who played tennis in her head subsequently improved from her seemingly vegetative state. Owen points out that by scanning her brain with fMRI, doctors were able to tell she was recovering long before she showed any outward physical signs. Early detection of a brain-damaged patient's potential for recuperation could lead to alternative treatments in the form of more aggressive drug or surgical interventions and to the encouragement of social interactions, such as visits from family members.

Owen's team is currently devising a protocol for "talking" to a vegetative patient's mind, by employing the same basic principles as in its initial test of the tennis player. "If the patient imagines playing tennis, it means 'yes.' If they imagine walking through the rooms of their home, it means 'no,'" Owen says. The different thoughts light up, or activate, various regions in the brain. With some practice on healthy subjects, the researchers have learned to tell apart thought-only responses of "yes" and "no" in under a minute. The doctors are now preparing to test their technique on a vegetative patient whom they have already found to be aware. If they succeed, they will "converse" with a locked-in person for the first time ever.

As with any new technology, it will take several years to understand how best to use fMRI in a clinical setting, and for now, researchers continue to deny most requests to scan brain-damaged patients. "It's not ethical, because we have not completed the research we would need to complete to be absolutely certain that our interpretations of the scan are right," says Hirsch, who fields e-mails from concerned family members on a daily basis. "It's just heart-wrenching, the number of people out there who want to know about the cognitive life of their loved ones who can't respond to them."

For now, using fMRI to diagnose or communicate with brain-damaged patients will continue to happen only in the small number of research laboratories devoted to studying disorders of consciousness. Funding is scarce for investigators studying brain damage, according to both Hirsch and Owen. The equipment is expensive--a state-of-the-art MRI scanner capable of functional scanning costs several million dollars--and scientists have not yet figured out the best way to use the relatively new technology. But with the possibility of being able to communicate with vegetative patients lurking just on the horizon, the researchers hope their work will eventually lead to the widespread release of locked-in minds.

"It's not something that every hospital can start doing yet," Owen says. "But we'd like to develop the technique so we can make it easier and accessible to everyone."