Technological advancement is growing at such a rapid pace that people with disabilities can now speak, walk, play games, and do much more using just their minds.
This development has been driven by progress in Brain-Computer Interface (BCI) technology. BCIs allow individuals to control external devices, such as computers, using only their thoughts. These interfaces work by acquiring brain signals, analyzing them, and translating them into commands that can be relayed to the output device.
BCIs have a wide range of applications, including assisting people with disabilities by restoring functions lost due to conditions like stroke, neurodegeneration, and spinal cord injury. By restoring capabilities such as vision, speech, and motor function, BCIs demonstrate the potential to enhance human abilities.
Depending on how the signals are taken from the brain, the Brain-Computer Interface can be termed either:
- Invasive, where electrodes are directly implanted in the cortex
- Semi-invasive, where electrodes are placed on the brain’s exposed surface
- Non-invasive, where electrodes are placed on the scalp
Over the past decade, research and development in BCI technology have grown significantly. Recent advancements have enabled individuals with severe paralysis to communicate by translating their brain signals into text and regain control over prosthetic limbs or their muscles. BCIs are also being explored to improve memory, enhance learning, and address mental health issues.
With so much going on in the field of BCI, let’s take a look at the most recent breakthroughs made by scientists that have the potential to significantly enhance human capabilities.
Restoring Speech Capability with High Accuracy
In a study published just this month, researchers developed a new BCI that translates brain signals into speech with an accuracy as high as 97%, which is one of the most accurate systems of its kind.
The sensors were implanted in the brain of Casey Harrell, a 45-year-old man who had severely impaired speech due to amyotrophic lateral sclerosis (ALS). Also known as Lou Gehrig’s disease, this neurological disorder affects the nerve cells controlling movement throughout the body and leads to a gradual loss of the ability to walk, stand, and speak.
When Harrell joined the BrainGate clinical trial, his speech was unintelligible, and others had to help interpret it for him. However, within minutes of activating the BCI system, he was able to communicate his intended speech.
UC Davis Health created this new BCI technology for people with neurological conditions who are unable to communicate via speech.
When a user attempts to talk, the new system interprets brain signals before turning them into text that is then ‘spoken’ aloud by the computer.
“Our BCI technology helped a man with paralysis to communicate with friends, families, and caregivers.”
– David Brandman, co-senior author and co-principal investigator of the study and a neurosurgeon at UC Davis
He also added that their experiment has demonstrated “the most accurate speech neuroprosthesis (device) ever reported.”
Brandman, who’s also an assistant professor in the UC Davis Department of Neurological Surgery and co-director of the UC Davis Neuroprosthetics Lab, implanted the BCI device in July last year. Four microelectrode arrays were implanted into the brain region responsible for speech coordination, the left precentral gyrus. These arrays utilize 256 cortical electrodes to record brain activity.
With this device, patients’ attempts to move muscles and talk are detected.
“We are recording from the part of the brain that’s trying to send these commands to the muscles. And we are basically listening into that, and we’re translating those patterns of brain activity into a phoneme – like a syllable or the unit of speech – and then the words they’re trying to say.”
– Neuroscientist Sergey Stavisky, the study’s co-principal investigator and an assistant professor in the Department of Neurological Surgery and co-director of the UC Davis Neuroprosthetics Lab
BCI technology has progressed greatly over the past many years, but efforts to enable communication have been slow and prone to frequent word errors. This is because machine-learning programs that interpret brain signals require a lot of time and data to perform.
The new system aims to remove that barrier to communication that makes it difficult for the user to be understood consistently.
“Our objective was to develop a system that empowered someone to be understood whenever they wanted to speak.”
– Brandman
The patient used the system in both spontaneous and prompted conversational settings. Speed decoding, in both cases, happened in real time with constant updates that kept it working accurately. The decoded words were then shown on a screen and read aloud by the computer.
The voice that read it aloud was actually the patient’s before he had ALS. For that, the researchers used software that was trained with existing audio samples of the patient’s pre-ALS voice.
The first speech data training session took the system just half an hour to achieve an impressive 99.6% word accuracy with a 50-word vocabulary. The next session took an additional 1.4 hours of training data to achieve a 90.2% word accuracy for a 125,000-word vocabulary. With continued data collection, the BCI maintained a 97.5% accuracy.
This ability to decode what the person with the implant is saying correctly about 97% of the time, according to Brandman, “is better than many commercially available smartphone applications that try to interpret a person’s voice.”
The study conducted 84 data collection sessions over 32 weeks. The patient used the BCI for communication via video chat and in person for more than 248 hours.
Enhancing Communication & Mobility with Unique BCI Systems
Another big development in the last couple of months came from Carnegie Mellon University, where researchers were able to achieve bidirectional BCI functionality. This achievement enhances the interface’s ability to interpret brain signals and send sensory feedback directly back to the brain.
Bin He, a professor of biomedical engineering at the University, and his group were successful in integrating focused ultrasound stimulation to realize dual communication for the first time.
In a study with 25 human subjects, the bidirectional BCI used machine learning to encode and decode brain waves. This development has the potential to significantly enhance the signal quality and overall performance of non-invasive BCI by stimulating targeted neural circuits.
While non-invasive BCI tech is safe and cheap, the fact that it records signals over the scalp instead of from inside the brain and the quality of its signals have limitations.
As such, the Carnegie Mellon University group has been working on improving the effectiveness of non-invasive BCIs. This has led to using deep learning approaches to decode just what one is thinking and then facilitating control of a robotic arm or cursor. Now, in their latest research, the group used focused ultrasound for precision non-invasive neuromodulation, and the findings showed a significant boost in the EEG-based BCI communication performance.
The human subjects in the study used a BCI speller, which is a visual motion aide commonly used by nonspeakers to communicate and spell out phrases like “Carnegie Mellon.”
In the study, the subjects put on an EEG cap and generated EEG signals just by looking at letters to spell out the words they wanted to.
By applying a focused ultrasound beam to the brain’s V5 area, which is part of the visual cortex, the researchers were able to improve the performance of the non-invasive BCI among subjects greatly.
“The BRAIN Initiative has supported more than 60 ultrasound projects since its inception. This unique application of non-invasive recording and modulation technologies expands the toolkit, with a potentially scalable impact on assisting people living with communication disabilities.”
– Dr. Grace Hwang, program director at the BRAIN Initiative at NIH
Yet another big development was seen last year when BCI enabled a paralyzed man, after suffering a spinal cord injury, to not just stand but also walk naturally. This brain-spine interface won the Physics World 2023 Breakthrough of the Year.
This system included two fully implantable devices, one recording brain activity related to leg movement and the other one electrically stimulating the spinal cord to control leg muscles. This created a ‘digital bridge,’ allowing an individual with paralysis of the arms and legs to walk.
The ECoG signals from the brain were monitored using a 64-channel electrode grid embedded in a titanium case the same thickness as the skull.
The technology was developed based on WIMAGINE, a unique implantable medical device able to record the brain activity at the cortex’s surface. A dedicated AI algorithm was also developed for real-time decoding of a patient’s intention of movement.
In the device’s clinical trial, a 38-year-old male with an incomplete cervical spinal cord injury from a bike accident a decade earlier had two devices surgically implanted in his brain and a paddle lead on his lumbar spinal cord.
Using the BSI, the patient was able to climb stairs, negotiate obstacles, and navigate changing terrains. Moreover, the BSI was stable and reliable for over a year of use without supervision.
But What About the Safety of BCI Devices?
Amidst all this progress, there have been many concerns regarding the handling of neural data and cybersecurity threats such as brain tapping and misleading stimuli attacks. Another big concern is regarding the safety profile of the brain-computer interface.
For invasive BCIs, the risk of brain infections, tissue damage, seizures, bleeding, and hemorrhage arises. In contrast, non-invasive BCIs may cause headaches, eye strain, and skin irritation from prolonged exposure to electromagnetic fields.
While these concerns are legitimate, a study by Brown University researchers analyzed close to two decades of safety data on clinical trials testing the BrainGate technology and found low rates of adverse events.
The BrainGate BCI has been in the making for more than twenty years, and the latest study concludes that its neurotechnology should continue to be evaluated for its potential to help people with paralysis turn thoughts about movement into actual action and regain lost neurologic function.
According to Dr. Leigh R. Hochberg, a brain science professor at Brown and a director of the BrainGate academic consortium leading the development and testing of the tech:
“In the largest ongoing trial of intracortical brain-computer interfaces, the interim safety profile reported today supports the possibility that these systems may become restorative neurotechnologies for people with paralysis.”
While intracortical BCI showcases huge potential in mobility and community restoration, Hochberg said for these advances to really translate into patient care, it all depends on “whether the devices are accompanied by an acceptably low degree of risk.”
The report on safety assessed just over 12,200 days of safety data spanning 14 clinical trial participants belonging to the 18-75 age group with quadriparesis resulting from ALS, spinal cord injury, or brainstem stroke.
Between 2004 and 2021, when all these patients were enrolled in the trials, 68 adverse events related to the device were found. Skin irritation was the most common issue, occurring in the small portion of the device.
Six adverse events were found to be associated with the surgical procedure. Two participants with a history of traumatic brain injury suffered brief seizures post-operation but were easily treated.
Despite their being adverse events, none of the documented ones were unanticipated. Also, none of these adverse events led to infections in the nervous system that required device removal or resulted in permanently increased disability.
While the reassuring safety findings of the BrainGate Neural Interface system have been a big step forward, much work remains, including ensuring that devices become fully implanted and are available to users 24/7.
Companies Making Strides in the Field of BCI
Synchron, Blackrock Neurotech, Kernel, and Emotiv are among the many organizations working to advance this technology and explore its applications in the medical, cognitive, and entertainment sectors.
The majority of these companies are privately held, and it’s rare to see public companies advance in this area. Those who are public are due to being part of larger organizations. For instance, NextMind, which developed a non-invasive BCI device, was acquired by Snap Inc. Another example is Ctrl-labs, a neural interface startup that was acquired by Meta (previously Facebook).
So, let’s take a look at some prominent names that are at the forefront of BCI technology:
#1. Neurable
Neurable is a neurotechnology company that develops AI-powered tools for brain signal translation and BCI technology. In May 2024, the company raised $13mln, which brings its total fundraising since its founding to over $30mln. In late 2019, Neurable raised $6mln in a Series A funding round to move beyond VR applications and develop non-invasive BCI, two years after it unveiled the world’s first mind-controlled VR game.
Now, with the fresh round of funding, the company aims to make its tech “accessible to everyone,” said CEO Dr. Ramses Alcaide, adding:
“We’re empowering individuals to understand their own mind, optimize human performance, and conquer the most pressing health challenges of our generation.”
#2. Precision Neuroscience
This company recently set a world record for the number of electrodes, 4,096 in total, placed on the human brain to record cortical data. The development came as part of Precision Neuroscience’s testing of its Layer 7 Cortical Interface. This achievement is expected to help the company “understand the brain in a much deeper way.”
Precision Neuroscience was founded by Benjamin Rapoport, a neurosurgeon who co-founded Neuralink but left, citing safety concerns.
#3. Neuralink
Founded by Elon Musk, Neuralink is a US-based private company developing high-bandwidth BCIs to treat neurological conditions and eventually enhance human cognitive abilities. The company’s focus is on creating implantable devices that can read and stimulate large numbers of neurons.
Most recently, the second person with a Neuralink brain chip, which, unlike the first one, remains fully attached to the brain, used the implant to play the popular video game Counter-Strike 2. The chip was planted last month, and the patient reportedly had a “smooth” recovery. The patient had lost control of limbs after a spinal cord injury.
This development occurred after the first patient had 85% of the electrodes attached to his brain displaced, although he could still use the implant effectively. The company reduced some mitigations to avoid the issue, and no thread retraction has been observed so far.
During his latest appearance on Lex Fridman’s podcast, Musk predicted that a human with a chip-enhanced brain defeating a professional video game player is not that far off now. Talking about the future, Musk also said that Neuralink’s long-term objective is to improve the symbiosis of AI and humans by enhancing an individual’s capacity to communicate at a large scale.
Conclusion
BCI technology has been gaining a lot of attention and momentum over the past few years, and this development shows no signs of slowing down. In fact, new records and breakthroughs are constantly being achieved that aim not just to allow people with disabilities to have more enriching lives but also show the potential to enhance capabilities in healthy humans.
While the road to implementing this technology is a long one for now, the continuous progress shows that what was once part of science fiction may soon become a reality.
Click here for a list of the best biotech stocks to watch.
