Brain-Computer Interface Allows Locked-In Paralyzed Patients To Communicate

Patients who are conscious, but completely paralyzed, have communicated with their doctors and families through a device that measures brain oxygenation. Surprisingly, the patients who have been given this opportunity apparently consider themselves happy. Although the technology is currently limited to yes-no answers, and not always accurate, it could enable significant improvements in quality of life, and lead to more advanced communication techniques.

Locked-in syndrome is literally a common nightmare. Sufferers can hear, feel, and sometimes see. It can occur suddenly through a stroke, or as the endpoint of progressive diseases such as amyotrophic lateral sclerosis (ALS), the condition that afflicts Stephen Hawking. Locked-in syndrome is not always 100 percent, and in famous cases such as Jean-Dominique Bauby, communication is possible through blinking.

Professor Niels Birbaumer, of the Wyss Center for Bio and Neuroengineering in Geneva, has gone a step further. In PLOS Biology he has described success in tracking the thoughts of patients with complete locked-in syndrome (CLIS) who cannot even control blinking. Birbaumer used a technology known as functional near-infrared spectroscopy (fNRIS) to measure the level of oxygenation in the frontocentral parts of the brain.

Birbaumer started by asking patients questions where he and the patients knew the answer, such as “Is Paris the capital of France?” and “Is Paris the capital of Germany?”. These two produced different oxygenation patterns, indicating that the patients were responding “yes” to one and “no” to the other.

From these questions, Birbaumer built up a pattern of how blood flow for each patient looked for positive and negative answers, before asking questions only the patient could answer.

Birbaumer was able to learn if the patients had immediate needs, and asked more general questions about happiness, with unexpected results.

CLIS is widely seen as a life of frustration and boredom, and many people express a preference for life support to be turned off should it happen to them. However, all four patients in this trial consistently indicated that they were happy over a period of weeks. This group may be an unrepresentative sample in this regard, as all four were sufferers of ALS, whose gradual deterioration forewarned them of CLIS, and each had given instructions to be kept alive.

Birbaumer's method remains imperfect. Patients got better at controlling the timing of their blood flow with practice, but Birbaumer moved to asking about their condition when the fNRIS gave the right answer to known questions 70 percent of the time. Although clearly better than chance, this is hardly accurate enough to make life or death decisions.

To compensate, Birbaumer asked important questions on multiple days, although the answers were sometimes disappointing. One family asked a locked-in father for approval of his daughter's marriage to her boyfriend, to which nine times out of 10 the reading indicated his opposition.