Most epilepsy-related Machine Learning (ML) tools tend to focus on seizure detection using readily available EEG "training" datasets. ML has potentially many other epilepsy-related use cases beyond seizure detection that are yet to be discovered, but the lack of "training" data is likely hindering its use. If tools could be developed to help identify and generate new "training" data from existing epilepsy datasets, ML will no doubt become more widespread in epilepsy research.
Idea No. 80