A Novel Segmented Peak Signal Method for Epileptic Seizure Detection Based on Frequency Band of Signal
Epileptic seizure detection is generally alludes to the utilization of methods to perceive that a seizure is happening or has happened. Automatic identification systems are used for this purpose and that is one such a time-consuming activity. Many automated systems are available in both machine learning and deep learning systems. Though machine learning methods perform efficiently, still there is a lot of time complexity. This time complexity may be reduced by deep learning methods but needs huge amount of data. In the previous proposal Peak Signal Features were extracted after the segmentation process and epilepsy was detected using machine learning algorithms. In our proposed Frequency Band Based Epileptic Seizure Detection, this amount of data is reduced by feature selection method which will perform manually using Peak Signal Values. In this paper, epilepsy is detected using the Deep Learning methods on frequency band of EEG segmented signals. The small variations of the frequency band of the brain wave signals will be analyzed for better performance. This will improve the overall performance and reduce the time complexity. The experimental result demonstrates the effectiveness of the proposed approach and it is very helpful in detecting seizures in minimum time.