In 2023, neurotechnology has witnessed remarkable progress in brain implants, which can record, decode, and even alter brain activity. This year’s developments in brain-machine interfaces have been fueled by the integration of artificial intelligence (AI), opening new possibilities for individuals with neurological disorders and paving the way for groundbreaking applications.
Restoration of mobility and memory
One of the standout achievements this year was the successful use of a spinal cord implant to translate the intentions of a patient with Parkinson’s disease. This condition gradually impairs movement planning.
By decoding the patient’s intentions, this innovative technology enabled the individual to regain the ability to walk freely after decades of limitation. This breakthrough holds promise for restoring movement in various other neurological disorders, such as Lou Gehrig’s disease and brain damage resulting from stroke.
Additionally, researchers conducted trials using electrical stimulation to enhance short-term memory in individuals with traumatic brain injuries. Timely electrical impulses increased attention spans in these patients, even many years after their injuries. This improvement allowed them to manage multiple daily tasks and engage in activities like reading, significantly improving their quality of life.
Brain implants have also proven to be valuable diagnostic tools. One study utilized implants to decode brain wave patterns associated with depression and potentially predict relapses. By discerning differences in brain signals between healthy individuals and those with depression, this research may lead to the development of more effective algorithms aimed at redirecting brain activity away from depressive states.
Transforming thoughts into text
One of the most notable achievements in 2023 is the development of technologies that can translate thoughts into words and sentences. This breakthrough is particularly beneficial for individuals who have lost the ability to speak due to neurological disorders or conditions such as stroke, paralysis, or locked-in syndrome.
Researchers at Stanford University achieved a remarkable feat by helping a 67-year-old woman afflicted by Lou Gehrig’s disease restore her speech at a rate of 62 words per minute—more than three times the speed of previous implants.
Their approach involved decoding speech-related electrical activity from Broca’s area, the brain’s language center, and the muscles surrounding the patient’s mouth. These signals were processed by a recurrent neural network (RNN), a deep learning algorithm, to identify the fundamental elements of speech. The system decoded the woman’s thoughts within three days, albeit with some errors.
Another system took a different approach, using a device known as electrocorticography (ECoG), which consists of small, plate-like electrodes placed on the brain’s surface to capture electrical signals. Although it requires surgical implantation, ECoG minimizes damage to sensitive brain tissues.
AI algorithms decoded neural activity associated with vocal movements, such as tongue and mouth positioning, while large language models constructed sentences from the data. This system translated brain signals into text at approximately 78 words per minute, albeit with a 25% error rate. To compensate for these errors, the implant utilized facial expressions to animate a digital avatar, providing patients with an additional mode of communication.
Wireless brain implants and non-invasive techniques
Brain implants have traditionally involved electrodes connected to cables linked to external computers for decoding neural activity. However, a significant development in 2023 introduced a wireless implant system. This system employed flexible, grain-sized circuit boards scattered across the brain to detect and temporarily store changes in activity.
These “nodes” wirelessly transmitted data to a receiver resembling headphones, which processed information, controlled brain stimulation through the nodes, and powered the array. Although implantation surgery is still required, this wireless approach represents a notable step forward in brain-machine interfaces.
Furthermore, researchers explored non-invasive methods of capturing brain signals. AI was employed to translate functional magnetic resonance imaging (fMRI) data into the essence of a person’s thoughts, capturing the evolving ideas even when the exact words were elusive. Another study utilized headgear embedded with electrodes that resembled a swim cap, placed on the scalp. As users silently read sentences in their minds, this cap, with the assistance of AI, translated their “thoughts” into text.
Some researchers even ventured into groundbreaking approaches involving light-based interfaces. A recent study combined genetically engineered neurons responsive to light with flexible probes with different colored LED lights.
With over a thousand independent LED pixels, this device could simultaneously control the activity of multiple individual neurons. This innovative technology helped distinguish specific brain circuits responsible for various mental functions, even deep within the brain.
Ethical considerations and future outlook
As brain implant technology evolves, it inevitably encounters ethical challenges. The prospect of a device translating thoughts into text raises concerns about privacy and boundaries. In response to these issues, UNESCO has released a blueprint outlining the need for global regulations and an ethical framework for neurotechnology.
This initiative aims to address the ethical implications of brain implants as they advance into uncharted territory.
