Related Experiment Video
Updated: Aug 12, 2025
06:04
Systematic Hearing Performance Evaluation Process for Adolescents with Cochlear Implantation at Early Ages
Published on: March 24, 2023
434
A high-performance speech neuroprosthesis
Biorxiv : the Preprint Server for Biology
|January 30, 2023
Summary
This study introduces a new speech brain-computer interface (BCI) that decodes attempted speech from brain signals. This advanced BCI offers faster, more accurate communication for individuals with paralysis, approaching natural conversation speeds.
Area of Science:
- Neuroscience
- Biomedical Engineering
- Assistive Technology
Background:
- Speech brain-computer interfaces (BCIs) aim to restore communication for individuals with paralysis.
- Existing speech BCIs lack the accuracy and speed for unconstrained communication from large vocabularies.
Approach:
- Developed a novel speech-to-text BCI utilizing high-resolution intracortical microelectrode array recordings of neural activity.
- The BCI decodes spiking activity associated with attempted speech movements.
Key Points:
- Achieved a 9.1% word error rate on a 50-word vocabulary and 23.8% on a 125,000-word vocabulary, significantly outperforming prior speech BCIs.
- Decoded speech at 62 words per minute, substantially faster than previous BCI records and approaching natural conversation.
- Identified key neural code features for speech, including spatially intermixed tuning and persistent articulatory representations, even years after paralysis onset.
Conclusions:
- This intracortical speech BCI demonstrates a viable method for restoring rapid communication in individuals unable to speak due to conditions like ALS.
- The findings pave the way for more effective assistive communication technologies.
