The scientist who helps the locked in speak again

Six years ago Elke Haenisch, now 70, was diagnosed with amyotrophic lateral sclerosis (ALS), a rapidly progressive incurable disease which attacks the body’s motor neurons which...
08 February 2017


Six years ago Elke Haenisch, now 70, was diagnosed with amyotrophic lateral sclerosis (ALS), a rapidly progressive incurable disease which attacks the body’s motor neurons which control movement, until the patient becomes totally paralyzed. In 2012, doctors diagnosed Haenisch as being completely locked-in. Haenisch may appear to be comatose but she is fully conscious. She cannot open or close her eyes without assistance, but she can hear and understand conversations. She can feel discomfort from lying in the same place for so long without being able to move a single muscle. She can be too hot or too cold, but while her family and doctors speak to her, she has been unable to alert anyone to how she really thinks and feels. Until now...

Niels Birbaumer is a 72 year old neuroscientist at the University of Tübingen who has devoted his life to the search for a way of communicating with patients in Haenisch’s condition. The eye muscles are the last to be affected by ALS and technological advances have allowed many patients in the latter stages of the disease, to operate assisted communication devices using eye movement. But even these muscles eventually fail, and for the patients who are completely locked-in, such devices are useless.

“The technology available means that with your eyes you can still do everything,” Birbaumer says. “If you lose the eye movement, then you can do nothing. The only possible chance of still communicating is through a brain-computer interface.”

Brain-computer interfaces (BCIs) which attempt to link neural responses to certain words or letters, have been under development for the past 30 years. However while BCIs often work when trialled in patients in earlier stages of the disease, their effectiveness has been limited for the completely locked in. “BCI's work fine with ALS patients who can still operate other communication devices using their eye muscles,” Birbaumer says. “But with the completely locked in, it has been difficult.”

In August 2014, Haenisch began to be treated by Birbaumer and his team at Tübingen. By this point, she had not been able to communicate with her husband and carer Michael for over two years.  Initially Birbaumer tried a BCI using a method called EEG to record her brain’s electrical activity along the scalp. He tried for a year, even implanting electrodes into her skull to try and get a stronger reading. There was no success.

“My theory is that if you are completely locked in for a long period of time, you lose volition or voluntary intention,” he says. “All the intentions, thoughts and wishes that you had, they all have no consequence anymore because nobody knows what you’re thinking. It’s like if you fall in love with a lady, but the lady never responds to you. What do you do? You give up. Without volition these BCIs cannot work because you need intention in order to select a letter.”

But Birbaumer had a solution. Instead of measuring brainwaves, he would measure brain metabolic changes, monitoring the blood flow in the frontal part of the brain using a technique called near infrared spectroscopy. Instead of trying to match changes in brain activity to letters, he would ask questions which required a simple yes/no answer.

“We look at the blood flow in the frontal part of the brain using laser light and at the same time we present simple sentences such as, ‘Is London the capital city of Germany?’ ‘Is London the capital city of England?’” he said. “Thinking yes or no occurs like a reflex. It happens automatically and doesn’t require strong volition.”

Birbaumer programmed a computer to ask Haenisch many hundreds of such questions, speaking in her husband’s voice. Slowly he began to get the responses he was hoping for. “The computer asks and then waits for 15 seconds,” he explains. “When it thinks it detects an answer, it says to the patient, ‘I think you said, ‘Yes/No’.’ And then when it thinks the patient is wrong, it asks, ‘Was that answer correct?’ Over time, you get fairly safe information about what’s correct and what’s not. Most patients can learn the system within a few weeks. I ask her whether she is at home and most of the time I hear ‘Yes’.’   

Birbaumer also realised that one of the keys with communicating with patients like Haenisch is being able to tell whether they are awake or asleep.

“When you’re in that state you go in and out of sleep because you have nothing to do and your sleep rhythm gets distorted,” he said. “There are days when you don’t get good responses because she’s sleeping a lot. But our system can now measure sleep, and we can detect when to question her and when to not.”

Once Birbaumer was confident he was obtaining the correct answers 70% of the time, he moved on to more serious, open questions:

“Are you in pain?”
“Do you want to see your husband?”
“Do you want to live?”

Like all completely locked in patients, Haenisch is kept alive by a series of machines. These feed her and help her breathe artificially. But while most medical professionals and family members often assume that the quality of life for people in this condition is terrible, this isn’t always true. “The interesting thing is that the patient’s family and doctors all assume that their quality of life in this state is very low,” Birbaumer says. “But when you ask the patient without the family or the doctor present, they say, ‘No I feel ok. I don’t have a problem.’” 

Birbaumer says there is one question in particular which predicts whether they want to remaining living or die:

“Do you think you are a burden to your family and to the world?”

“If the answer is ‘Yes,’ then when they become completely locked in, they don’t want to go on artificial respiration and they die,” Birbaumer says. “This happens for 90% of these patients.”
Birbaumer says the patients who desire to continue living almost always have a very positive family environment. “That makes a big difference. When you are completely paralyzed, your attention focuses on what you have. If you only get positive stimulation all day long, you feel that everybody loves you and you feel very good. And there’s another factor. We’ve found that complete paralysis leads to a kind of complete relaxation. Almost like meditation. The brain senses that the muscles are not working anymore and responding to its signals and so it interprets this as a state of deep relaxation.”

Haenisch is unwavering in her desire to live. But now a line of communication has been established, Birbaumer has been facing legal battles to maintain it. 

“The machinery costs 50,000-60,000 euros and the insurance companies don’t want to pay this. Right now we have many lawsuits going on. We have won the first one already and I am hopeful we will win the rest. There are very few patients who are completely locked in so in total, the cost is not really very high. With this system I now have 6 patients who can all use it to communicate.”

So far Birbaumer’s work has only been applied to patients who have become locked in as a result of ALS. But he believes the same technique can be used to detect whether patients paralyzed as a result of accidents, are capable of communicating. Currently 30% of patients are misdiagnosed as being in a vegetative state when they are actually locked in.  Researchers in Cambridge, Liege and Ontario have made progress in determining whether these patients are consciously aware or not but unlike Birbaumer, they have had no success so far in communication.

As ever, the problem is cost. “It’s taken us thirty years to make this work. But it’s extremely time consuming, and with the normal research grants that are available, it’s difficult to finance this kind of research. In the German system it’s easier as people are more patient and you get more long-term money.”

Birbaumer hopes that it will be eventually possible to use his system to establish more advanced conversation with these patients, rather than a simple yes/no response.

“To do that we would need to be sure we’re getting the correct answers at least 80% of the time,” he says. “Then we can use the brain response to select letters. But this is difficult. The problem is that the brain is a noisy organ. While the patient is thinking their answer, there are millions of processes happening simultaneously and I have to extract this one simple thought.”

Birbaumer accepts this next stage may be a problem for future generations to solve. For now, he is working on making his system practical so the families of his patients can have basic communication on a daily basis.  

“We need to prove that this system works in the long-term and it can be reliably operated by the families themselves. Then I will be happy. My work will be done and I can die myself.”


Add a comment