assistivetechnologyblog.com
Brain Computer Interface Breakthrough: Mind-to-Word Technology Achieves Record Word Decoding Rates, Paving the Way for Enhanced Communication
In a groundbreaking development, two separate research teams from Stanford University and the University of California, San Francisco, have unveiled remarkable advancements in brain-computer interface (BCI) technology that decode words from a person’s thoughts, specifically targeting individuals who have lost their ability to speak. These BCIs represent a significant leap forward in the field of assistive technology. The Stanford University team, led by Jaimie Henderson, implanted electrode arrays into the cortex of an ALS patient referred to as “T12” to record neural activity. They then employed a deep-learning model to translate this activity into words, achieving an impressive word decoding rate of 62 words per minute (wpm). This rate is over three times faster than the previous record of 18 wpm, set by the same Stanford research group in their efforts to decode handwriting from neural activity. Meanwhile, the UC San Francisco team, under the guidance of Edward Chang, took a different approach. They used electrocorticogram (ECoG) electrodes placed on the brain’s surface to capture neural signals related to vocal tract movements, such as lip, tongue, and jaw movements. Their BCI system achieved an even faster word decoding rate of 78 wpm, surpassing the Stanford team’s device and setting a [...]
Venkat