Losing your voice or the ability to type a simple message can mean, for someone who suffers from a severe paralysisThis is much more than a physical limitation: it involves seeing their autonomy, relationships, and, to a large extent, their very life plans diminished. In recent years, neurotechnology has focused on this problem, seeking ways for these individuals to communicate again without relying on slow and exhausting systems.
In this context, a US research team has successfully tested a implantable brain-computer interface capable of transforming finger movements into text on a virtual keyboard. The device, still in the experimental phase, has been tested on two people with almost complete paralysis and has achieved typing speeds and accuracy levels close to those of a person without motor disabilities.
A neuroprosthesis that translates typing attempts into letters
The work is signed by scientists from Institute of Neuroscience Mass General Brigham, in Boston, and of the Brown Universitywho have been collaborating for years in the BrainGate consortium, an initiative focused on developing brain-computer interfaces for people with paralysisThe new study, published in the scientific journal Nature Neuroscience, describes a writing neuroprosthesis that does not move a cursor, but instead relies on a conventional QWERTY keyboard as a starting point.
To achieve this, the researchers implant microelectrode sensors in the motor cortexThe region of the brain involved in controlling voluntary hand and finger movements. These microelectrodes detect the electrical activity that occurs when a person mentally attempts to move their fingers to press a key, even though the body cannot execute the gesture due to the injury.
In front of the participant is shown a standard QWERTY keyboard accompanied by a schematic representation of the fingers. Each letter is associated with a specific combination of finger positions (for example, up, down, or flexed). When the user imagines these movements, the electrodes collect the neural signal and send it to a computer system that... translates into text characters.
The process doesn't stop there: the decoder output goes through a predictive language modelSimilar to the autocorrect feature on mobile phones, it helps to correct errors and complete words, so that the final sentence is coherent and as faithful as possible to the patient's intention.
Two patients with severe paralysis as a test case
The trial was conducted with two people with very advanced paralysisOne participant with advanced amyotrophic lateral sclerosis (ALS) and another with a cervical spinal cord injury which caused him to be quadriplegic. Both were part of the BrainGate clinical program and gave their consent to test the new writing neuroprosthesis.
Following the microelectrode implantation surgery, the volunteers underwent brief training with the system. It took approximately 30 calibration phrases so that the software could adjust its decoding algorithms to each person's neural signals. From there, they were asked to write messages using only their attempts to move their fingers on the virtual keyboard displayed on the screen.
The results were remarkable for their speed and accuracy. One of the participants reached a top speed of 110 characters per minuteThis translates to approximately 22 words per minute, with a word error rate of 1,6%. That margin of error is similar to that of a person typing by hand on a physical keyboard or on a smartphone screen.
The second volunteer, suffering from advanced ALS and requiring mechanical ventilation, also succeeded produce understandable sentences through the system, although at a somewhat slower pace. In his case, the significance of the progress is especially noteworthy, since he had completely lost the ability to speak and could not use conventional assistive technologies without enormous effort.
One particularly important aspect of the trial is that both patients were able to use the device in your own homeAnd not only in a strictly hospital or laboratory setting. This suggests that, with further development, the technology could be integrated into everyday support systems so that those suffering from severe paralysis can communicate from home with family members, caregivers, or healthcare professionals.
Why this interface is different from current systems
Today, many people with paralysis who retain some eye control rely on eye-tracking devicesThese systems allow users to select letters or icons by moving their eyes across a screen, but, as described by the patients themselves, they are slow, tiring to use, and prone to errors. In many cases, users end up abandoning them due to the frustration they cause.
BrainGate's neuroprosthesis takes a different approach: instead of tracking gaze or moving a cursor with thought, it focuses on decode the finger movement attempts on a keyboard that is familiar to almost any literate person. This strategy has two clear advantages: on the one hand, it allows us to take advantage of the motor memory that many patients developed over years of using physical keyboards; on the other, it facilitates reaching higher write speeds than those of other augmentative communication systems.
Furthermore, the use of artificial intelligence algorithms, both for decoding the neural signal and for the language model, contributes to improved accuracy without requiring excessive cognitive effort from the user. The person doesn't have to "think of individual letters," but rather imagine moving their fingers as they would if they were actually typing.
According to the research team, this combination of implantable sensors, advanced signal processing, and language models transforms the brain-computer interfaces in an increasingly solid alternative to existing solutions, at least for a specific group of patients with severe paralysis who do not find an adequate response in conventional systems.
The role of the BrainGate consortium and future projections
The development of this neuroprosthesis is part of the work of the consortium BrainGateThe International Neuroscience Association (INCAA), founded in 2004, brings together neurologists, neuroscientists, engineers, computer scientists, neurosurgeons, mathematicians, and other specialists from various academic institutions. Their common goal is to create technologies that enable... recover lost functions in people with neurological diseases, spinal cord injuries, or amputations.
Over the past two decades, BrainGate has demonstrated in controlled trials that brain-computer interfaces can be used to control cursors, robotic arms, or external devices based on brain activity. The now published advance focuses specifically on written communication, a key area for those who have lost both the ability to speak and the ability to use a physical keyboard.
Those in charge of the trial emphasize that the technology is still in the research phase. Questions such as the following remain to be resolved: durability of implants, the stability of the signals over time, the possible risks associated with surgery, the ease of use of domestic systems or their fit into the financing of health systems, including European ones.
Even with these precautions, the team believes that the device opens a path for the industry to develop in the medium term. commercial versions of neuroprostheses adapted to patients with different paralysis profiles. The consortium emphasizes that collaboration between academic centers and companies will be key to translating these experimental results into real clinical solutions.
Relevance for patients in Europe and upcoming challenges
Although the study was conducted in the United States, its impact is directly relevant to people with ALS, spinal cord injuries or stroke In Europe, including Spain, where population aging and the increase in neurodegenerative diseases raise the demand for communication assistance technologies year after year.
In healthcare systems like Spain's, with a strong public component, these types of developments are usually evaluated not only for their clinical effectiveness, but also for their cost-effectiveness and its ability to integrate in neurological rehabilitation networks and home care. The possibility of brain-computer interfaces functioning in home environments favors their future adoption, provided that the equipment is simplified and the components become cheaper.
Looking ahead to the coming years, researchers are proposing several areas for improvement. One of them involves introducing custom keyboards or stenography systems that allow for even faster typing; another, to take advantage of the same technology to try to restore reaching and grasping movements in people with upper limb paralysis, using the already identified patterns of neural activity.
There is also talk of combining these interfaces with other support tools, such as screen readers, voice assistants, or home automation devices, with the aim of building more accessible living environments for those who depend on a wheelchair or require continuous care. All of this with the same common thread: reconnecting the person's brain with the world around them when their body has stopped responding.
Taken together, this new clinical trial demonstrates that a implantable brain-computer interface It can give people with severe paralysis a form of written communication that is fast, accurate, and stable enough to be used in everyday life, which is an important step towards solutions that not only prolong life but also allow people to live with greater autonomy and ability to relate to others.