From Science to Technology: The Case of the Locked In

In February 2010, the New England Journal of Medicine published a stunning article.   Monti, NEJM Researchers in groups led by Stephen Laureys, at the University of Liege, and Adrian Owen, then at Cambridge, reported the results of functional magnetic resonance imaging (fMRI) experiments they conducted with 54 people diagnosed as in minimally conscious or persistent vegetative states.  In spite of being completely unresponsive, four of the patients showed brain activations similar to those of healthy people when told to do two tasks – to imagine playing tennis or to imagine navigating through their home (or the streets of a familiar city).  One of those four patients, one who had been diagnosed for five years as being in a vegetative state, was able to answer five “yes” or “no” questions, such as “Was your father’s name Alexander,” by making the appropriate response – imagining tennis for “yes,” imagine navigating for “no” (or vice versa).

The finding was fantastic, and frightening. At least some people diagnosed as vegetative had normal brain reactions – and at least one, the only one they tested, could answer questions and was apparently conscious.  Consciousness without being able to respond, not even to blink an eye – is that solitary confinement, hell, or both?

But along with the terror came hope.  That apparently vegetative, but actually “locked in,” patient could, in fact, communicate.  But – this was the quite serious catch ­– only with an MRI.

Time in an MRI is neither cheap nor easy.  Moving apparently vegetative patients to a scanner can be complicated.  Involuntary movements in the scanner by the patients can ruin the scan.  Patients with implanted metal devices cannot be in an MRI.  And, even if one were able to use fMRI to determine which apparently vegetative or minimally conscious patients were, in fact, completely or largely locked in, the patients could scarcely be left, full time, in a scanner with a healthy person standing ready to ask “yes” and “no” questions.

The 2010 paper was an important proof of principle, but it could do little to help the hundreds, or thousands, or tens of thousands of people diagnosed as vegetative or minimally conscious but actually locked in.  The science was there, but not the technology to use it effectively.

That seems to have changed this week with a publication in The Lancet by the same group.  Cruse, Lancet (registration required)  For this article, the group did a very similar experiment with 16 patients at Liege or Cambridge, all, this time, diagnosed as in a vegetative state, five as a result of traumatic injury and eleven with non-traumatic injury (mainly hypoxic injury, produced by a lack of oxygen).  But instead of using expensive and difficult fMRI, they used a lower tech, more flexible, more portable, and much, much cheaper technology – electroencephalography (EEG).  This time 3 of their 16 patients were able to follow commands.

The patients were fitted with an EEG cap and then asked to do the following:

“Every time you hear a beep, try to imagine that you are squeezing your right-hand into a fist and then relaxing it . . . . Concentrate on the way your muscles would feel if you were really performing this movement. Try to do this as soon as you hear each beep”

Sometimes, instead of being told to make a fist with their right hands, they were told instead to wiggle all the toes on both their feet, then relax.  The researchers then analyzed, with sophisticated statistical methods, the EEG signals from their patients to see whether any of their EEG results had changed significantly when asked to perform these tasks of imagining.

For three of the 16, the researchers saw significant results.  These were not perfect results – the patients’ brains responded between 61 percent and 78 percent of the time.  But the brains of the other patients did not respond significantly at all.  Two of the five patients with traumatic injuries responded; only one of the eleven with non-traumatic injuries did.  (In the fMRI experiment, the only responders had traumatic injuries; the majority of those patients also had non-traumatic injuries but none of them gave any results.  Interestingly, only nine of the 12 healthy controls returned meaningful signals when asked to do the same task.

This research did not include the equivalent of the one patient who answered “yes” and “no” questions in the earlier study.  The researchers seem not to have tried that – yet.  It is, presumably, possible that their brain responses were automatic and not associated with consciousness.  The authors discuss the possibility and consider it highly unlikely, but it would be encouraging to see the next step taken.  The fact that a quarter of healthy controls could not activate the EEG is also somewhat disconcerting – if a quarter of healthy brains could not generate a distinctive signal, how many of the patients had similarly had working brains that returned no signal?

This experiment is certainly not the end, nor can it fairly be called the beginning.  The fMRI experiments get that honor.  But it is the beginning of what may prove a useful technology.  EEGs are cheap, they are not very sensitive to motion artifacts, they can be used in patients with metal implants, and they can be done at the patient’s bedside.  More work needs to be done, but the path seems clear to an EEG-based method of both detecting consciousness in those apparently vegetative or minimally conscious people and giving them access to regular, effective, and economically plausible communication.

That would, indeed, be a victory for these people with disabilities and for the people who love them.  And it is another reminder of the importantly different roles of science and technology.  Science can tell us that something can be done, but, to change lives, and especially to change the world, we need not just science but technologies that make that science a real possibility.  As well as economic, social, and legal systems that make it an available possibility.

So, kudos to Laureys, Owen, and their team. Good work – now get back to it, with whatever help you may need from the rest of us.

Hank Greely