Diagnosis of active epilepsy in resource-poor setting
Prediction models based on EEG characteristics
R package developed to extract features from multivariate time series from EEG data and feed them into a random forest classifier.
As this was a short lasting project, we mainly focused on processing multi-variate EEG (Electroencephalography) data as a proof of concept.
The software handles the data format and structure used in one particular study carried out by colleagues from Utrecht University in data collected in sub-saharan Africa, but uses generic external libraries (caret) for the machine learning.
This software was developed as part of the Young eScience Award 2015 (awarded to Wim Otte in 2015, but project took place in 2016). The project description can be found here https://www.esciencecenter.nl/project/diagnosis-of-active-epilepsy-in-resource-poor-setting A technical report on the project can be found here:https://www.biorxiv.org/content/10.1101/324954v1