Brain-computer interfaces

Computer interfaces

Fighting disabilities

Communication, User Interfaces, Motion Control
for people with locked-in syndrome (ALS, sclerosis multiplex, brainstem stroke etc.)

Science-fiction

Transport safety (Boeing), game interfaces, military; later: everyday user interfaces (Neuralink), symbiosis with AI, brain-brain interface

Variability of electric potential on the scalp (up to 100 µV) due to brain activity

Can detect evoked and induced potentials – brain waves responding to stimuli

People can direct their attention or imagine things → non-invasive BCI!

(from: "Positive, 300µs after stimulus")

Event-related potential, evoked when the brain receives an expected stimulus

Task of the user: count hearts on top of the answer they want

Receiving the answer: see if P300 signals coincide with hearts being shown on one of the aswers (average over a few hearts for accuracy)

User imagines movement of a hand or a leg

Neuronal oscillations induced in the sensory-motor area (~10 Hz)

Advantages

Disadvantages

requires training and repeated imagining of the same motion
a lot of electrodes needed to help localise signals to the left/right hemisphere

Seeing an oscillating light evokes a brain wave with the same frequency (+ higher harmonics)

5 Hz – 35 Hz

Strong response, easy to detect

Annoying, exhausting for eyes

Risk of inducing an epileptic seizure

35 Hz – 75 Hz

Light flickering not visible

SSVEP are still detectable, but weaker

Special screen, options illuminated by oscillating LEDs with different frequencies

Task of the user: just focus on the chosen answer

Receiving the answer: find which frequency is strongest in the EEG signal

Brain-computer interfaces allow receiving information from the brain directly (without muscle intermediaries)
They make it possible people with locked-in syndrome communicate with the outside world
Non-invasive BCI methods are based on EEG
Three most common paradigms are P300, Motion imagery, and SSVEP