Epilepsy Detection Through EEG Signal Classification Using Machine Learning Methods

In this exertion demonstrates the application of a deep learning algorithm to the analysis of focal and non-focal EEG data for the finding of epilepsy. This research presents a strategy for automating the diagnosis of epilepsy by making use of a soft computing resource. By using the ANFIS classification technique, focal and non-focal samples may be distinguished. Structures computation, features classification, and features transformation are the three pillars upon which this technique rests. For an electroencephalogram (EEG) data, the Approximate and Directional bands are produced when the suggested NSCT is used. NSDF and Non-Subsampled Pyramid Filter are the two main components of this NSCT. These fall into two categories: low pass, which includes power frequencies, and high pass, which includes high frequency noise. To do this, we construct directed subbands and reroute the high-pass signal using NSDF. Step two of decomposition involves creating high and low frequency sub bands by passing the low pass sub band signal via the Non-Sub Sampled Pixel Frequency (NSPF). The directed sub bands that are created are achieved by retransmitting the high-frequency sub band signal via NSDF. Using the statistical properties obtained from these sub-bands, the ANFIS classification method divides the overall EEG signal into focal and non-focal samples. Applying the proposed method to a collection of EEG signal extraction measurements allows for validation.