Research Paper
Brain-to-text decoder achieves 62 words per minute in paralyzed patients
PubMed3/30/2026·03/30/26🌐 USA
Summary
Stanford trained a decoder on intracortical recordings from 3 ALS and spinal cord injury patients, decoding attempted handwriting at 62 WPM with 94% accuracy using a 26-letter vocabulary plus punctuation and an online correction interface.
Why it matters
At 62 WPM with high accuracy, this system exceeds most augmentative communication devices and approaches natural typing pace, making it genuinely useful for locked-in patients today.
#BCI#Neural Interfaces#Neuroimaging#Neurostimulation
Want the full NeurotechMag Pro intelligence database?
Access 800+ papers, trials, patents, funding rounds and news items — updated daily.
Start free trialAlready a member? Sign in