The experiment was conducted on a 36-year-old man who was paralyzed after a stroke damaged his brainstem at age 20. The participant is cognitively healthy and able to make out grunts, but the paralysis prevents him from articulating words, meaning that he is suffering from complete anarthria.
In his day-to-day life, the participant communicated by slightly moving his head and typing on a special device, but this device doesn’t allow him to communicate very fast - his typing speed is about 5 words per minute. With the new technology introduced to the participant, he was able to produce 15 words per minute (with an error rate of just 26%).
The sensorimotor cortex (highlighted in purple and green in the table above) is the part of the brain responsible for one’s ability to articulate words by moving the lips, tongue, and throat. Over the course of 48 training sessions, the participant was taught to think about saying 50 specific words presented to him on a screen. The computer picks up the brainwaves from the sensorimotor cortex and uses something the authors call “deep-learning algorithms” to recognize words.
To speed up the process of communication further, the computer also has a “natural-language model,” which deduces the next word based on previously-uttered words. All of this happens in real-time, which allows the participant to make immediate requests like “I need my glasses” and “I am thirsty.”
“While the intervention is quite invasive, requiring brain surgery to implant a recording strip on the surface of the brain, and the ‘thought-to-spoken’ conversion accuracy was modest, the paradigm is groundbreaking,” said Dr. Lee H. Schwamm of the American Stroke Association.
To give you more perspective on the significance of this technology, we must add that this is the first device of its kind to be used in a patient who is not able to speak. Previous devices were only tested in patients who could speak. It is also the first technology able to produce immediate results, which makes it more practical. The next step in this research could be investigating the technology’s potential in patients with aphasia, a more common language deficit that occurs after a head injury or stroke.
Share this information with family and friends!
