Open-source tool picks the right brain signals for mind-reading devices

Researchers built BCI-sift, a Python toolbox that automatically identifies which brain signals matter most for decoding tasks like speech. Tested on implanted electrodes recording from speech areas, the tool consistently picked the same informative channels across different people and improved classification accuracy by filtering out noise. The same approach works for any brain-interface device, potentially speeding up clinical BCI development.