A brain-machine interface (BMI) developed by neuroscientists allows primates to navigate a robotic wheelchair with nothing but their thoughts.
The interface may prove to be a ray of hope for people whose disabilities have cost them most muscle control and mobility due to ALS or quadriplegia.
The basic principle of operation involves the use of the BMI to use signals from hundreds of neurons recorded simultaneously in two brain areas of the monkey responsible for sensation and movement.
The computers then translate the brain activity that occurs when the monkey thinks about moving towards a bowl containing fresh grapes, into real-time operation of the wheelchair.
This technology moves past the drawback of the use of non-invasive measures, such as an EEG, for mind control of robotic devices. This is because there are severely disabled individuals, to which actions as basic as blinking is an impossible task.
As such, a non-invasive procedure may not be sufficient. The use of intracranial implants has clearly shown to offer better control of a wheelchair than with the use of non-invasive devices.
Experiments for this technology began in 2012, led by Miguel Nicolelis, M.D., Ph.D, and co-director for the Duke Center for Neuroengineering.
To start, the scientists implanted hundreds of hair-thin microfilaments in the somatosensory and premotor brain regions of two rhesus macaques. The scientists then trained the macaques to navigate the wheelchair toward a bowl containing grapes.
As the training phase progressed, the scientists recorded the large-scale electrical brain activity of the primate. Using this data, the researchers embarked on programming a computer system to translate brain signals into digital motor commands that controlled the movements of the wheelchair.
Continued learning made the rhesus macaques eventually get better over time in terms of efficiency and timely completion of the trials.
The team also made a groundbreaking discovery during the trials. They noticed that the brain signals of the monkey showed signs that the monkeys were contemplating their distance to the bowl of grapes.
This signal was in addition to those that translational and rotational movement, but interestingly was not present at the beginning of the training. It seemed to have emerged thanks to the growing proficiency of the primates to perform the task.
This discovery thus provided new insights into the enormous flexibility of the brain to assimilate a device and correlate the spatial relationships of the device to the surrounding world.
Going forward (no pun intended), the team hopes to expand the experiment by recording more neuron signals. Currently, they measure nearly 300 neurons in each monkey. In the near future, they expect to record up to 2,000 neurons using the same technique.
The increased coverage of neurons will increase the fidelity and accuracy of the primate BMI before the team seeks trials for an implanted device in humans.
The researchers published the details of their experiment in the journal Scientific Reports.