Brain-Computer Interface for Locked-in ALS Patients

Brain–computer interface (BCI) technology promises to allow patients with amyotrophic lateral sclerosis (ALS) who have lost the capacity to communicate, even with their eyes, to "talk" to their loved ones, or at least to respond to questions.

A new study, published online January 31 in PLoS Biology, showed that over several weeks, patients with ALS in a complete locked-in state (CLIS) using the BCI technology had over a 70% correct response rate to questions posed.

The technology uses functional near-infrared spectroscopy and electroencephalography to measure blood oxygenation and electrical activity in the brain.

If replicated, the positive results could indicate "the first step towards abolition of complete locked-in states, at least for ALS," said Niels Birbaumer, a neuroscientist at the Wyss Center for Bio and Neuroengineering, Geneva, Switzerland, who at the time of the study was at the Institute of Medical Psychology and Behavioral Neurobiology, University of Tübingen, Germany.

Until now, communication has been next to impossible for those in a CLIS who have total motor paralysis but intact cognitive and emotional processing.

Many attempts at neuroelectric BCI communication have failed in these patients with CLIS, said the authors.

As ALS progresses, it leads to respiratory difficulties and often to bulbar dysfunction, which affects muscles of the face, head, and neck. Most patients eventually lose control of the last muscular response, usually the eye muscles, rendering them completely "locked-in."

The study included four patients with ALS: a 68-year-old woman with bulbar ALS who was in a CLIS; a 76-year-old woman, also with bulbar ALS and in CLIS; a 61-year-old man with nonbulbar ALS in CLIS; and a 24-year-old woman with rapidly progressing juvenile ALS and on the verge of CLIS.

All had lost reliable eye communication and all were being cared for at home.

The CLIS state in these patients was verified by the attending neurologist, electrooculography recordings (measures of the corneo-retinal standing potential between the front and the back of the eye), and video recordings of the families' failures to achieve contact with the patient.

The authors noted that all patients were enrolled in the study consecutively.

Computer Recognition

While hooked up to a machine that measures frontocentral oxygenation changes, the patients were asked "yes" and "no" questions, the answers to which they and their relatives knew the correct answers (eg, "You were born in Berlin" or "You were born in Paris").

After asking hundreds of such questions over time, a computer recognizes what is a "yes" brain response and what is a "no" brain response for a particular patient.

"The computer calculates the average answers to those hundreds of yes or no questions," explained Dr Birbaumer.

Three patients completed more than 46 sessions over several weeks, and one patient completed 20 sessions.

Communicating with the frontocentral cortical oxygenation-based BCI had an above-chance-level correct response rate over several weeks. "In 70% of these questions, the computer has reliably recognized the difference," said Dr Birbaumer.

It didn't appear to matter how long it had been since a patient had become completely locked in; one patient had been locked in since 2009 but the rest, more recently. However, Dr Birbaumer stressed that because there were only four patients, "we can't talk about statistics."

The researchers also asked the patients "open" or "general attitude" questions, such as "Is your life wonderful?" or "Is your life terrible?"

For such questions, it's assumed that a patient would have the same response when asked at different times. "Whether I ask the question today or at the end of the week or next week or a month later, usually your answer would be stable, so it's relatively unlikely you would answer differently."

But because there are no right or wrong answers to such questions, it's difficult to validate responses to them, although researchers tried to confirm the information with caregivers and family members.

To illustrate how useful information gleaned with the BCI technology can be, Dr Birbaumer used the example of one of the patients, who developed an open back wound while being transported in a wheelchair. When asked if she was in pain, her brain said "yes" 100% of the time, and then later, when the wound was healed, she responded "no" 100% of the time when asked if she was in pain.

Life Satisfaction

Interestingly, the patients generally indicated that they were satisfied with their quality of life. "Uniformly, the data — not just from our laboratory, but from laboratories from all over the world — show that quality of life in these locked in people is high," said Dr Birbaumer.

This tends to be true except for immediately after diagnosis (when patients can become suicidal) and when they lose the capacity to breathe on their own.

Dr Birbaumer believes that their quality of life is relatively high because they appreciate what they still have and have lost "goal-directed thinking."

"They have stopped thinking about the things they know they can't achieve."

He also pointed out that the patients included in the study were being cared for at home — by loving families with healthcare system supports — and had accepted artificial ventilation when breathing on their own had become impossible.

"This means that these patients have already decided for themselves that they want to live; so in that sense, we already have a positive selection," he said. "It's clear that a positive family relationship is a crucial factor and without that, people die."

Although there's no good documentation, an estimated 95% of patients with CLIS die "under mysterious circumstances," noted Dr Birbaumer. This could be through withholding of medication or means of respiration.

How was he sure that the patients in the study weren't cognitively affected by their disease, as they had lost all capacity to communicate using motor function?

"You can argue that these questions are fairly primitive and not indicating that they are cognitively intact," said Dr Birbaumer. "Neurologists tell me over and over again that these patients have brain damage, but the evidence for that is not very high."

He also stressed that "we asked hundreds of knowledge questions; we asked hundreds of geographic questions; and then we asked hundreds of personal questions. If these people answered correctly to these questions, that means they understood the questions."

The authors noted that electroencephalographic oscillations and electrooculographic signals did not exceed the chance-level threshold for correct communication despite occasional differences between the physiologic signals representing a "yes" or "no" response.

Families of the patients in the study experienced "substantial relief" with use of the system, the authors report. The husband of one of the patients in the study is already using the BCI machine at home and the families of the others are learning to use it, said Dr Birbaumer.

Most families of patients with ALS don't come forward for technological help until they're "desperate," said Dr Birbaumer. However, he added, most neurologists don't know about BCI, so they don't recommend it for these patients.

Speller System

Although encouraged by the new results, Dr Birbaumer would like to build a "speller system" whereby patients with ALS could mentally select letters and words for even better communication. This would involve inserting invasive electrodes and has previously not been very effective.

However, with improved technology and surgical techniques, the hope is that locked-in patients with ALS will be able to use brain responses with such a system to "say what they want," said Dr Birbaumer.

In a press release, John Donoghue, director of the Wyss Center, said his center plans to build on the results of this new study to develop clinically useful technology for people with paralysis resulting not only from ALS but also from stroke or spinal cord injury.

For a comment, Medscape Medical News approached Nick Ramsey, PhD, from the Brain Center Rudolf Magnus, University Medical Center Utrecht, the Netherlands.

Dr Ramsey has published a paper on implanted BCI technology that has allowed a locked-in patient with ALS to wirelessly transmit her thoughts to a computer to spell out messages on a tablet.

Although he believes that implantable devices are "the way to go" for patients with ALS, Dr Ramsey called the efforts of Dr Birbaumer and his colleagues to find a way for those who are completely locked-in to communicate at the bedside "commendable."

"What this group has shown is that you can see changes in blood flow that indicate that people still have an intentional brain activity."

However, the 70% successful response rate "is not particularly good," said Dr Ramsey. "And you have to ask the question multiple times before you have an answer, so it's a minimal way of communicating — but it's more than nothing."

Dr Ramsey believes patients with ALS should be offered an implantable device before their brain loses the ability to generate an intention.

In commenting on the relatively high life satisfaction study patients expressed, Dr Ramsey agreed that as their disease progresses, people tend to adjust to what they can do and don't focus on what they can't do.

"I don't think it's a conscious choice; I think it's almost a biological thing where your mind adjusts to what it can do."

Patients with ALS "still have a lot of pleasure to spread; they just have a smaller range of possibilities of where to spread it to," he added. "The small things become more significant."

Dr Ramsey reported and he and his own team are considering more candidates for their neuroelectrical BCI and are working to improve the software to make it more useful.

The authors have disclosed no relevant financial relationships.

PLoS Biol. Published online January 31, 2017. Full text

For more Medscape Neurology news, join us on Facebook and Twitter

Medscape Medical News © 2017  WebMD, LLC

Send comments and news tips to news@medscape.net.

Cite this: Brain-Computer Interface for Locked-in ALS Patients - Medscape - Feb 01, 2017.