Tag: real-time-speech

1 article

Industry News

Paradromics's first Connexus participant used a brain implant to speak her own words

Paradromics announced on 14 September 2026 that the first participant in its Connect-One Early Feasibility Study, a Michigan woman in her 60s with progressive motor neuron disease and severe dysarthria, has used the Connexus intracortical brain-computer interface to generate her own words, answer open-ended questions, and hold a live phone call with her grandchildren, three months after her June 2026 implant at University of Michigan Health. Verbatim first sentence: 'Okay, I have a lot to say.' Second sentence: 'Don't be afraid. It's a wonderful experience to be able to help others by your participation in the research study.' Both were decoded in real time, word for word, with no errors, according to Paradromics. Output modality is text plus synthesized speech (not avatar-mediated). Quantitative benchmarks (words per minute, word-error rate, vocabulary size, decoding latency) were not disclosed. The Connect-One EFS runs across three sites (University of Michigan Health, UC Davis Health, Mass General Brigham) under FDA IDE approval granted in November 2025. Same-day academic milestone at UCSF (Edward Chang team, Nature Neuroscience) demonstrates simultaneous speech and gesture decoding via animated avatar. The two results are categorically different (commercial vs academic, text-plus-audio vs avatar-mediated) and should not be conflated.

Sep 16