Restoring Communication: A Brain-to-Text System


By: Inas Essa

A new study has revived the hope of restoring the ability to communicate through handwriting in people paralyzed from the neck down through converting thoughts into text with 94.1%accuracy.

The brain-to-text system converts thoughts to text, allowing people with severe paralysis or locked-in syndrome—which may be caused by brain stem stroke, traumatic brain injury, or other diseases—to communicate more efficiently through performing thought dictation.

Researchers at Stanford University worked on combining machine learning and a Brain-Computer Interface (BCI) to read visualized letters in the brain and display them on a screen with the help of two implanted electrode arrays that record the brain’s electrical activity. After that, this formulated text is gathered and processed by a computer that converts it into words on a screen; a form of “mental handwriting”, which means no hands are needed in the process.

The system has been tested on a 65-year-old man paralyzed from the neck down due to a spinal cord injury in 2007. In 2016, researchers implanted two small BCI chips into the patient’s brain to record the neuronal activity of the brain as he imagines writing each letter. The chips were implanted in the motor cortex that controls the movement of the arms and hands, allowing the researchers to profile brain-activity patterns associated with written language.

Restoring Communication

Although a previous technique in this field had helped analyze the neural patterns associated with speech and could decode imagined arm movements, it only allowed people with paralysis to type around 40 characters per minute, which is lower than the average keyboard typing speed of around 190 characters per minute. In this study, the patient succeeded in mentally typing 90 characters per minute, which is not so far from the average for most texters on a phone: 115 characters per minute.

Restoring the typing ability has been challenging since each letter triggers a different pattern of neural activity and generating speech is much faster than the typing or writing ability. As such, reclaiming people’s ability to communicate at the speed of their thoughts has been considered such a leap in this field.

Although it is hard to predict when this technology would be implemented into a real device, researchers hope that one-day millions who are severely paralyzed and unable to type or speak because of impaired limbs or vocal muscles would benefit from it and restore communication to express themselves and speak their thoughts out loud.

 


References