Motor-Imagery Brain-Machine Interface: C3/C4 Traces and Decoder Error (2D)
A subject with live synthetic C3 and C4 traces, the decoded cursor and the two overlapping log-power distributions with the decision threshold. The same shared model and numerical checks as the 3D scene. Educational, not clinical.
An educational model of how an EEG brain-machine interface tells left-hand from right-hand imagery. Imagining a movement lowers 8-12 Hz mu power over the opposite motor cortex; a decoder compares the log band power of C3 and C4. Change ERD depth, mu strength, window length and threshold in a Blender-built EEG cap scene, and read exact accuracy, d prime and bits per minute, with Monte Carlo and closed-form checks built in. Not a clinical tool.
A subject with live synthetic C3 and C4 traces, the decoded cursor and the two overlapping log-power distributions with the decision threshold. The same shared model and numerical checks as the 3D scene. Educational, not clinical.
2D · HTML5 Canvas 2D · 60 FPS target · runs fully client-side, no install