Internet of Medical Things based EEG Epileptic Seizure Detection

Seizures can range in severity from moderate to severe and are frequently unpredictable due to aberrant electrical activity inthe brain. In this paper, authors have designed a Internet of Medical Things based system for the Classification of Epileptic Seizure Detection on EEG Signals. This paper tends to the requirement for cutting-edge seizure identification techniques utilizing EEG signals using different pre-processing techniques like smoothening, and sharpening. The different parameters like Hjorth, and statistical features were extracted which were selected using the Meta Heuristic approach, a particle swarm optimization (PSO). The features were classified using different machine learning techniques. The paper incorporates a thorough examination of accuracy results with and without the Meta-heuristic methodology. The proposed model accomplishes a 96% accuracy utilizing SVM for linear kernel and PSO. The significance of each component of the dataset outlines EEG data, demographic information, and clinical history.