Check out our deep learning approach for the identification and onset prediction of electrographic seizure patterns (ESPs) in brain data recorded with the responsive neurostimulation device. iESPnet is a convolutional and recurrent neural network able to predict the presence of ESPs and their onset across patients and data cohort with an accuracy of about 90% and time onset error in between 3.6 and 5 seconds.
More information can be read here: https://onlinelibrary.wiley.com/doi/abs/10.1111/epi.17666
The code is open and available at GitHub: https://github.com/Brain-Modulation-Lab/Paper_iESPnet
Comments