9th International Conference on ICT for Sustainable Development

Authors - Arshpreet Kaur, Kumar Shashvat
Abstract - Identifying inter-ictal activity in EEG amidst of artefacts is a primary challenge for diagnosis of epilepsy. Artifacts corrupt the EEG; such that identifying inter-ictal activity becomes difficult. There are different artefacts that present the EEG some of the prominent ones are movement artifact, eye movement, electrode and emg artifact. The evaluation of EEG is a subjective procedure and detecting inter-ictal activity among the artifacts is a challenge. We have designed convolution neural network to identify between different artifacts and inter-ictal activity. For this work five sets have been considered, the first four sets compare each artefact with inter-ictal activity and set E, compares all artefacts with inter-ictal activity. The proposed method has achieved 100% accuracy in identifying electrode and emg artifacts from each other while for other artefacts the method also performed efficiently.