Glossary

neural decoding

Using software, increasingly machine learning, to turn recorded brain signals into information: an intended movement, a word, an image, or an estimate of activity the device cannot directly measure.

Decoding is what makes a brain–computer interface useful. It is also where the questions of cognitive liberty begin, because the better the decoding, the more a recording can reveal.

Documented real-world example: in the UC San Diego study behind transparent brain-computer interfaces, a neural network called a BiLSTM (bidirectional long short-term memory), which is built to handle signals that change over time, learned to predict deep calcium imaging activity in mice from surface electrical recordings alone. The authors reported it could predict both average activity and the activity of single cells.

Sources: Neural decoding — Wikipedia; Transparent Brain-Computer Interface Uses AI and Nanotech — Psychology Today (Cami Rosso, 24 Jan 2024)

Previous Next