Intracortical BCI advances speech decoding with prosody/loudness control
Key Questions
What advances in speech decoding were reported by the Chang lab?
The peer-reviewed study in Nature Communications showed vPCG encoding of speech mode and loudness at 94% and 89% accuracy in closed-loop tests with paralyzed participants. This builds on prior intended speech decoding work for severe dysarthria and ALS, with a recent report claiming 97% thoughts-to-text accuracy.
What does the research indicate about brain signals after paralysis?
Persistent signals were observed post-paralysis, supporting ongoing BCI viability. Sensory integration ties were also noted as relevant to speech processing alongside motor aspects.
How do speech memories relate to sensation versus motor control?
Related findings indicate that speech memories depend more on sensation and sound than on motor control, aligning with auditory neuroscience perspectives on BCI speech applications.
Chang lab Nature Comm peer-reviewed vPCG encoding of speech mode/loudness (94%/89% acc, closed-loop) in paralyzed trial participants, building on intended speech decoding for severe dysarthria/ALS; recent report claims 97% thoughts-to-text accuracy. Persistent signals noted post-paralysis + sensory integration ties. Pending full ITR/bpm, stability/safety, cohort/training details.