EEG Signal Classification using Deep Transfer Learning Technique in an Internet of Medical Things Environment

Successful categorization of epileptic electroencephalogram (EEG) signals is a crucial part in making a correct diagnosis of epilepsy. The findings of an electroencephalogram (EEG) test are also used in this categorization to assess whether or not the EEG signal represents an epileptic signal (healthy). There have been many suggested automated algorithms to help neurologists with the classification of these signals. Although these methods have often resulted in high average accuracy during classification, there is still room for advancement in terms of performance. We thus present a multi-segment EEG-based convolutional neural network for epilepsy-related EEG data classification. To train and test the neural network convolutional thoroughly, the spectrogram picture is fed into it. Each EEG signal segment on each test channel is first classified using the convolutional neural network model based on the training results, and then the results are combined in an ensemble for a final classification.