A neural prosthetic has been used for the first time to directly restore walking movement to the legs of nonhuman primates. A wireless “brain-spinal interface” was used by an international team of scientists to restore intentional walking movement to a temporarily paralyzed leg of a pair of rhesus macaques. The interface bypasses spinal cord injuries.
David Borton, an assistant professor of engineering at Brown and one of the study’s co-lead authors, explained that the system uses signals that have been recorded from the motor cortex of the brain. These signals then trigger the coordinated electrical stimulation of the nerves in the spine responsible for locomotion. The monkeys used in the study had nearly normal locomotion while the system is switched on. Borton added that evidence suggests that a spinal stimulation system controlled by the brain may enhance rehabilitation after a spinal cord injury. This study is a further step toward testing that hypothesis.
The research could aid in the development of similar systems to help humans with spinal cord injuries.
The collaboration was led by Grégoire Courtine, a professor at EPFL. He has started clinical trials to test the spine component of the interface in Switzerland. He warns that it may take a number of years before all the components of this intervention can be tested in humans, as there are numerous challenges that still lie ahead.
A complex interplay among neurons in the spinal cord and brain makes walking possible. Electrical signals that start in the brain’s motor cortex move down to the lumbar region in the lower spinal cord. There they activate motor neurons to coordinate the movement of the muscles that are responsible for flexing and extending the leg.
An injury to the upper spine often severs communication between the lower spinal cord and the brain. Although the spinal neurons and the motor cortex may both be fully functional, they are not able to coordinate their activity. The objective of the study was to re-establish some, or all of that missing communication.
A pill sized electrode array is implanted in the brain and is used by the brain spinal interface. The array records signals from the motor cortex. Ongoing pilot clinical trials using the technology was previously done in a study led by neuro engineer Leigh Hochberg from Brown. In those trials, people with tetraplegia controlled a robotic arm simply by thinking about the movement of their own hand.
Borton was part of a team that developed a wireless neuro sensor in the neuro engineering lab of Brown professor Arto Nurmikko. The sensor sends the signals gathered by the brain chip to a computer wirelessly.
After the signals are decoded, they are sent back to an electrical spinal stimulator wirelessly. The stimulator is implanted below the area of injury in the lumbar spine. The electrical stimulation is delivered in patterns that are coordinated by the decoded brain, and these then signal the spinal nerves responsible for locomotion.
The researchers implanted the wireless transmitter and brain sensor in healthy macaques to calibrate the decoding of brain signals. The signals that were relayed by the sensor were then mapped onto the animals’ leg movements. The team demonstrated that the decoder was able to predict the brain states associated with the flexing and extending of leg muscles accurately.
Borton noted that the transmitting of brain signals wirelessly was critical to this work. Freedom of movement would be limited with brain sensing systems that are wired. This would in turn limit the information about locomotion researchers could gather. Borton added that they were able to map the neural activity during natural behavior and in normal contexts, due to them being able to transmit the brain signals wirelessly. Such untethered recording technologies will be critical if neuro prosthetics are someday to be deployed to help human patients during daily life activities.
Courtine’s lab at EPFL previously developed spinal maps and these were combined with the researchers’ understanding of how brain signals influence locomotion. The spinal maps pinpointed the neural hotspots in the spine responsible for locomotor control and enabled the team to identify the appropriate neural circuits that had to be stimulated by the spinal implant.
With these components in place, the researchers tested the complete system on two macaques that had lesions spanning half the spinal cord in their thoracic spine. The researchers noted that macaques with this type of injury normally regain functional control of the affected leg within approximately a month. The system was tested in the weeks following the injury, while the macaques did not have volitional control over the affected leg yet.
During the test, the monkeys moved their legs spontaneously while walking on a treadmill with the system turned on. Kinematic comparisons with healthy control animals showed that the injured macaques were nearly able to achieve normal locomotor patterns with the assistance of the brain-controlled stimulation.
Although proving that the system works in a nonhuman primate is an important step, much more work needs to be done before the system can be tested in humans. The researchers also pointed out several limitations in the study, one of which is the inability of the system to return sensory information to the brain. Although it was clear that the limb was able to bear some weight, researchers were not able to test how much pressure the animals were able to apply to the affected leg.
Borton feels that they need to do more quantification in future full translational studies to determine how balanced the animal is during walking. They also want to measure the forces the monkeys are able to apply. He adds that the research sets the stage for future studies in primates despite the limitations.
He is hopeful that the technology will be used as a rehabilitation aid in humans at some point in the future.
Research has been published in the journal Nature.