WaveComputing

The brain runs on waves. Every EEG signal carries a record of them — slow rhythms, sharp transients, rolling and rotating patterns moving across the cortex. Clinicians have read these waves for a century; the waves themselves, what they are and how they combine, remain the real subject.

WaveComputing offers a language for interpreting brain signals and a model for the cortical processes that generate them. The same language extends naturally to other oscillatory signals whose information lies in waveform shape and harmonic structure.

EEG analysis is a mature field with excellent tools. The aim here is not to compete with them, but to give their measurements physical meaning — connecting what the signal is to what the brain does.

The work currently lives on the simplest closed surface that carries waves: a ring. Wave dynamics on the ring give a vocabulary for waveform shape, harmonic structure, and synchrony, and forward simulations of the substrate generate predictions for measurements made on real EEG by conventional methods. Seizure initiation, propagation, and termination can be reinterpreted in this language as transitions in the wave field — seizures as catastrophic collapses of the brain's capacity to sustain rich wave patterns. A spherical surface, two-dimensional and far richer, is the longer-term destination.

The framework's geometric structure — closed surfaces with continuous symmetries — is shared by other domains in which oscillatory signals encode meaning through their shape, and the same substrate can be applied there.

The project builds on three decades at the intersection of clinical epileptology, EEG research, and brain-inspired computing.

Kaspar A. Schindler, Sleep-Wake-Epilepsy-Center, Inselspital Bern