Predicting seizures with an 'army of algorithms'

How crowdsourcing can work to help individual patients

Five years ago the world’s first in-human trial of implanted seizure prediction devices was deemed a success.


Using implanted electrodes, the devices measured brain activity to calculate seizure risk in real time and warned patients via an external light that shone blue (low risk), white (medium risk) or red (high risk).

It was revolutionary for some patients, predicting seizures with 100% sensitivity, according to results in Lancet