Training MEMS devices for forecasting epileptic-seizures

Epilepsy is a chronic neurological disorder of the central nervous system that causes sudden loss of consciousness and convulsive seizures. A third of worldwide 50 million patients’ condition don’t benefit from existing drugs. We aim to implement physical reservoir-computing on a micro-resonator device and ‘train’ it to identify ‘pre-seizure’ patterns in EEG signal to acquire the ability of predicting seizure-events ahead of time.