Brain-Controlled Exoskeleton Aids Walking in Tetraplegia

Damian McNamara

October 09, 2019

A 28-year-old man with limited arm movement following a severe cervical spine injury can now move his arms in multiple directions and walk using an exoskeleton suspended from the ceiling.


The bilateral implantation of wireless recorders above the sensorimotor cortices of his brain and the decoding of epidural electrocorticography data that were generated enabled these accomplishments.


The patient relearned how to move his arms and legs using a brain-computer interface, first actively engaging in computer games, then moving a computer avatar, and ultimately graduating to the exoskeleton.

"For the first time, a tetraplegic patient was able to walk and control both arms using this neuroprosthetic, which records, transmits, and decodes brain signals in real time to control an exoskeleton," study author Guillaume Charvet, project chief of the Brain Computer Interface project at Clinatec in Grenoble, France, told Medscape Medical News.

He is one participant in an ongoing feasibility study of the strategy. The 2-year outcomes of the study were published online October 3 in Lancet Neurology.

"This project, named Brain Computer Interface, intends to prove that, after training, a person with serious motor capable of controlling complex functional substitution devices, such as a four-limb exoskeleton, by decoding his brain electrical activity," Charvet said.

This is not the first study in which an exoskeleton was used to promote rehabilitation or neurologic recovery after a stroke or severe spinal cord injury. However, previous work focused on hand orthotics or exoskeletons that were placed around an arm or the lower limbs. Charvet and colleagues took a more whole-body approach.

In past research, control of many movements was achieved using wire microelectrode recordings, although "a clinically compatible solution to compensate for motor deficits still does not exist," Charvet and colleagues note.

The current investigators designed the WIMAGINE device. It is fully implantable and biocompatible so as to allow long-term use.

To determine the best placement of each implant with its 64 electrodes, the researchers asked the patient to make real or imagined movements with his arms and legs. He repeated these movements while investigators imaged his brain using magnetoencephalography (MEG) and fMRI. This imaging allowed them to identify the center of his sensorimotor cortex for precise surgical implantation.

Voluntary and Virtual Movements

The patient's tetraplegia resulted from a C4–C5 spinal cord injury. He could contract his biceps to move his arms at the elbow and was able to move his left wrist. Otherwise, his sensory motor deficit was complete.

At study entry, the only assistive technology that he used was a wheelchair, which was controlled by a left arm support joystick. He was a candidate for the study because his fMRI and MEG recordings showed that cortical signals were produced when he imagined himself moving all his limbs.

The patient performed progressively more challenging tasks. Examples include a video game similar to Pong, in which he mentally controlled a paddle to intercept a falling ball. He also reached out to touch a target as part of a two-dimensional game and used his fingertips to interact with a three-dimensional panel of eight light-emitting diodes.

The patient's decoded epidural electrocorticography (ECoG) brain signals allowed him to control the 65-kg (143-pound) exoskeleton.

Reaching Milestones

The patient achieved 54% of hits in 19 experiments with the Pong-like video game task using his left hand.

When performing two-dimensional tasks with his left hand, he was successful 80% of the time. With his right hand, he was successful 82% of the time.

When using the avatar on screen to complete his tasks, he completed three-dimensional tasks 57% of the time with his left hand and 53% of the time with his right hand.

In reach-and-touch tasks using the exoskeleton, the patient successfully completed three-dimensional tasks with his left hand 69% of the time and with his right hand 62% of the time.

In initial walking tasks performed using the video game, the patient achieved a true positive rate of 83% and a false positive rate of 13%. When instructed to walk using the avatar, he achieved a true positive rate of 92% and a false positive rate of 5% per minute.

When wearing the suspended exoskeleton, his true positive rate was 73% and his false positive rate was 7% 2 months after surgery. The patient is shown walking in this video. He covered a total distance of 145 meters (480 steps in 39 periods of walking).

The patient also performed multi-limb activation of the exoskeleton to generate models that simultaneously controlled several degrees of freedom in combined tasks. For example, he was 84% successful performing a two-dimensional, two-handed task and 71% successful in completing a three-dimensional, two-handed task.

A Number of Firsts

The implant is the only device approved for a long-term clinical trial, enabling long-term recordings on a large number of channels ― 64 ― which is the highest number of contacts reported per side, Charvet said.

"This is the first demonstration of a high-dimensional control of bimanual neuroprosthetics allowing eight degrees of freedom, using a neural population recording system that is safe and compatible with chronic clinical use," he added.

The study is also the first to bring together all the elements needed for long-term clinical use in humans, including epidural recording, wireless power and emission, online decoding of many ECoG channels, and being totally embedded, the researchers note. "Our intervention showed no signal degradation, no side effects, and long-lasting tolerance," they write.

The 28-year-old patient in the study was the second person in whom the system was implanted. An initial participant was excluded because the recorders stopped communicating shortly after implantation, and the devices were removed. Investigators identified the technical problem and corrected it before implanting the system in the second patient.

"Our patient already considers his rapidly increasing prosthetic mobility to be rewarding," the researchers note. However, there is a caveat. "This progress has not changed his clinical status. The main goal of this report is to show that bilateral, semi-invasive, epidural chronic implants that control an exoskeleton with four limbs move us closer to achieving the expected progress in the field of deficit compensation."

"Three other tetraplegic patients will be included in this clinical trial in the coming years. This proof of concept will open the door to new applications for use of this neuroprosthetics at home by patients in their everyday lives," Charvet said.

"The Clinatec team is working on integrating new effectors, such as a wheelchair, and developing even robust and more precise algorithms to perform more complex movements, with the hope of later enabling tasks such object manipulation," he added.

Is Conscious Control Necessary?

"An originality of this study is showing the control of four limbs, whereas in most previous studies only one limb was controlled. However, autonomous walking with equilibrium is not so far possible," Tom Shakespeare, PhD, of the Department of Clinical Research at the London School of Hygiene and Tropical Medicine, United Kingdom, writes in an accompanying editorial comment.

"Although this study presents a welcome and exciting advance, we must remember that proof of concept is a long way from usable clinical possibility. A danger of hype always exists in this field.

"People with tetraplegia do already have usable solutions, in the form of lightweight wheelchairs with new generation batteries, and controls that can enable users to drive themselves by blow and sip, by micromovements of one hand and by other means," he notes. "Although this study suggests the possibility of replacing a joystick by conscious control, why this is a major practical improvement is not obvious."

Although a newly paralyzed patient "does indeed dream of walking again, a person who has adapted to their situation might have other priorities — eg, bladder or bowel management, pain control or avoidance of pressure sores," Shakespeare writes. "Indeed, people with spinal cord injury generally enjoy a good quality of life, regardless of the level and degree of lesion.

"Understanding the life goals of this patient group would be an important step towards collaborating on a genuinely useful medical or technological advance," he adds.

The study was funded by the French Atomic Energy Commission, the French Ministry of Health, the Edmond J Safra Philanthropic Foundation, Fondation Motrice, Fondation Nanosciences, Institut Carnot, and Fonds de Dotation Clinatec. Charvet and Shakespeare have disclosed no relevant financial relationships.

Lancet Neurol. Published online October 3, 2019. Abstract, Comment

Follow Damian McNamara on Twitter: @MedReporter. For more Medscape Neurology news, join us on Facebook and Twitter.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.