Neural NoiseMotor imagery is the mental rehearsal of a movement without overt execution. When a person imagines moving a limb, characteristic changes occur in sensorimotor cortex rhythms — particularly event-related desynchronization in the mu (8-12 Hz) and beta (13-30 Hz) frequency bands — that can be detected by eeg and used as control signals for bci-and-neural-decoding.
Motor imagery-based BCIs are among the most widely studied non-invasive BCI paradigms. Users learn to modulate sensorimotor rhythms through imagined hand, foot, or tongue movements, with classifiers distinguishing between different imagery classes to generate device commands. This approach requires no external stimulation and relies on endogenous brain signals, making it suitable for patients with severe motor impairment including locked-in syndrome.
Challenges in motor imagery BCI include substantial inter-subject variability in the ability to produce classifiable patterns (sometimes termed “BCI illiteracy”), the need for extended training, and relatively low information transfer rates compared to other BCI paradigms. Research directions include adaptive classifiers, transfer learning across sessions and subjects, neurofeedback-guided training, and hybrid approaches that combine motor imagery with other signal modalities to improve performance and usability.