新型脑机接口系统可实时将想法转化为语言

Core Insights - A brain-computer interface system developed by American researchers utilizes artificial intelligence to decode brain activity of users attempting to speak, aiding individuals with severe speech impairments to communicate effectively [1] Group 1: Research Overview - The study was led by researchers from the University of California, Davis, involving a 45-year-old male participant who lost clear speech ability due to amyotrophic lateral sclerosis (ALS) [1] - The system implanted 256 microelectrodes in the participant's brain to capture relevant signals every 10 milliseconds, allowing for real-time speech generation [1] - The system can convey pitch variations and emphasize chosen words, enabling the participant to hum notes in three different pitches [1] Group 2: Technological Advancements - Unlike earlier models that required 3 seconds to output speech or only produced speech after complete sentence imitation, the new system can replicate the speaker's voice and produce speech within 10 milliseconds of detecting neural signals [1] - The system not only conveys the speaker's intent but also captures natural speech features such as tone, pitch, and stress [1] Group 3: Challenges Ahead - Despite the breakthroughs, the widespread application of brain-computer interface technology faces challenges, including risks associated with electrode implantation surgery and the need for further validation of device stability and durability [2] - Protecting user privacy and data security remains a critical issue that needs to be addressed in the future [2]