‘Plug-and-Play’ Control Brain Computer Interfaces Have Arrived

Researchers develop the first self-learning BCI platform that doesn’t require daily recalibration.

By Simon Spichak | 6 October 2020

(Photo by Michael Dziedzic on Unsplash)

Physicist Stephen Hawking suffered from a disease that progressively impaired his ability to move. This disease, called amyotrophic lateral sclerosis, kills off the specific neuronal cells that help coordinate movement. For decades he depended on caregivers, a computer for communication and his wheelchair.

In the face of these circumstances, he persisted and continued contributing to our knowledge and understanding of physics. Imagine if Hawking, and others like him, had access to technology that allowed for seamless communication with a computer.

Every year, millions of people suffer motor impairments as a result of accidents, strokes, or neurodegenerative disorders. There is a tremendous mental health impact as these people struggle with a lack of independence and difficulty with every-day tasks. Unfortunately, we lack disease-modifying treatments for many of these disorders. There’s an enormous burden on the affected individual as well as their family and caregivers.

Researchers who are interested in neuroprosthetics look towards augmenting these impairments. Even a small incremental improvement could immensely benefit millions of people. By mapping our brain activity, we could connect other technology and control it with our thoughts. These brain-computer interfaces (BCIs) are entering the very early stages of development. Early-stage clinical trials aim to augment motor ability and rehabilitation.

It’s no surprise that medical-device companies developing these interfaces, like Neuralink and Cortera Neurotechnology, capture interest from the general public as well as investors.

By 2027, the market size for BCIs is estimated to reach $3.85 billion. Scientists from the Department of Neurology at the University of California, San Francisco (UCSF) demonstrated a plug-and-play interface that allowed a paralyzed (tetraplegic) individual to control a computer cursor on a computer screen. They published their findings in Nature Biotechnology on September 7th, 2020.

To fully appreciate the impact of this technology, let’s look at how we initiate movement.

A Primer On Neuroanatomy


Our brain moves by activating specific networks of cells, called neurons. Neurons possess long processes called axons that propagate electrical signals like conducting wires. Once an electrical signal reaches the end of an axon, the cell releases a neurotransmitter that signals with the next neuron on this circuit. These axons require insulation to prevent the signal from degrading as it travels. This insulation is called a myelin sheath.

When we’re infants, we practice and learn how to activate and use these networks to coordinate our voluntary motor movements. These networks become consolidated over-time, making us much better at controlling our limbs.

GIF generated by Opus Design

The motor cortex

The brain itself holds specific representations of different regions of our body. When we imagine the brain, we think of a grey, wrinkled structure. The wrinkled outer-structures are portions of the cerebral cortex, which is key to interpreting our senses and coordinating voluntary movement.

The motor cortex, which coordinates our voluntary movements, is a neural representation of our entire body. Beginning this endeavor in the 1940s, neurosurgeon Wilder Penfield mapped this cortex. While performing surgery on epileptic patients, he found that electrical stimulation could induce the perception of specific senses (i.e. the smell of burnt toast) or movements. Similarly, he also mapped the sensory cortex.

By providing small electrical stimulation, looking for corresponding movements in his patients, he mapped the motor cortex. The motor cortex is organized in a somatotopic manner, meaning that specific parts of this cortex corresponded to regions of the body. These representations didn’t correspond 1:1 to body size.

What we’d look like if our motor cortex representation directly mapped onto our bodies. | Dr. Joe Kiff (http://psychology.wikia.com/wiki/User:Lifeartist / http://psychology.wikia.com/wiki/User:Dr_Joe_Kiff) / CC BY-SA

Augmenting motor and communication impairments

These networks are learned and developed from a lifetime of use and movement. If the regions of the brain involved in moving a certain limb or any of the downstream motor neurons are damaged, it leads to severe motor impairments. When the damage is quite severe, we cannot consolidate a new pathway to circumvent the damaged path.

Since we understand the pathways that our brain uses to move our limbs, we can use these controls to modify many different disease states.

With motor neuron or Parkinson’s disease, cells required to move specific parts of the body die. In a stroke, specific cells in the motor cortex could become damaged. Locked-in syndrome is particularly troubling as an individual retains all of their senses but cannot move and struggles to communicate.

These individuals lose their sense of agency because their speech may become impaired. Right now, these technologies could help encode new pathways to control a computer output. Something as simple as moving a mouse cursor to select letters or messages is immensely helpful. In the far future, neuroprosthetic augmentation could directly interface with the brain, restoring motion in these limbs.

BCIs allow us to generate these new pathways using the brains’ own flexibility. Through the process of neuroplasticity, we could interface with a computer to move or control prosthetic augmentations. However, the current state of technology requires daily re-calibration.

The new ‘plug-and-play’ interface

ECoG Array | Source: Noah Berger / UCSF

Imagine playing a video game on your PC. Every day you start the game, you must figure out the controls through trial and error. Such mechanisms make it difficult to build and strengthen skills over time. Imagine if it took 30 minutes just to figure out the controls every single day. This is the extremely frustrating reality representing the current state of BCI enhancements and prosthetics. Daily recalibration takes time and prevents continual learning and improvement.

A plug-and-play interface would allow the participant to use the device immediately after turning it on. Rather than relearning the controls every single day, they learn and build their skills with the device.

Karunesh Ganguly MD, Ph.D., senior author of a recent study that tested a “plug and play” brain prosthesis, described this problem:

“The BCI field has made great progress in recent years, but because existing systems have had to be reset and recalibrated each day, they haven’t been able to tap into the brain’s natural learning processes. It’s like asking someone to learn to ride a bike over and over again from scratch.”

Other studies already show the ability of BCI recordings to allow paralyzed individuals to point, click, and select letters on a screen. It’s not entirely clear why, but these implants display instability for the day. The electrodes implanted within the brain may shift slightly, affecting the signal received by the BCI.

If the consistent signal begins to shift, performance degrades. Individual neuronal recordings don’t remain stable with these devices, making it impossible to assign individual electrical signals from cells to specific controls or functions.

In the “plug and play” brain prosthesis study, one participant was paralyzed and had a chronic electroencephalograph (ECoG) implant. This measured the activity of brain waves at 128 sites at once providing spatial resolution. While the participant could not move their limbs, they could still think about the movement, activating motor areas in the brain. By imagining neck and wrist movements while staring at the screen, this individual learned to use the BCI to control the mouse cursor which generalized to other tasks on a computer screen. While the brain encoded the ideal pathway for completing these tasks, the algorithm also consolidated a memory of this pathway — a first in this area of research.

The patient is also enrolled in a separate study that aims to augment movement with neuroprostheses. These self-learning algorithms incrementally progress this field forward to clinical applications. Ganguly spoke of the potential impact of this work:

We’ve always been mindful of the need to design technology that doesn’t end up in a drawer, so to speak, but which will actually improve the day-to-day lives of paralyzed patients.

Additionally, this solution would allow for daily-use of neuroprosthetics. Since the individual wouldn’t have to relearn the controls every day, they might learn to live with these prosthetic augmentations!

Final thoughts

There are currently 35 ongoing or active clinical trials involving BCIs. These interfaces have a wide range of applications. For spinal-cord injuries, stroke, and degenerative disorders, these interfaces may restore or augment motor function. Allowing patients to speak again or perform more tasks independently will also markedly improve their psychological well-being.

While it may be years (or decades) before we perfect neuroprosthetic technology, the BCIs in this study are extremely useful. Even controlling the text or mouse cursor on a computer screen is immensely valuable for individuals with locked-in syndrome. Though unable to move, they remain completely aware of their environment.

To further advance this technology, brain implantations must become less invasive. Technologies like neural dust show promise as small, wireless devices with single-cell resolution. It’s unclear, however, when these devices will advance towards clinical trials and testing in humans.

By combining the incredible learning abilities of the human brain and computer algorithms, researchers developed a plug-and-play BCI system. I look forward to seeing these BCIs enhance the quality of life for millions of people around the world with motor and communication impairments. This may be the first step towards fully restoring function.

Reprinted with permission from the author.

Simon Spichak is Co-Founder at Resolvve Inc. He is a neuroscientist, science communicator and frequent writer. You can subscribe to his newsletter at simonspichak.substack.com. Follow him at Medium and Twitter.

Brain Gate – Breakthrough in Brain-To-Computer Interfaces

Thought control of robotic arms using the BrainGate system

Providing a Sense of Touch through a Brain-Machine Interface

Be sure to ‘like’ us on Facebook


Please enter your comment!
Please enter your name here