Received: 13.05.2021; Revised: 09.07.2021, Accepted: 30.07.2021, Published Online: 31.08.2021

Ashok Vajravelu

Faculty of Electrical and Electronic Engineering, Universiti Tun Hussein Onn Malaysia, Malaysia. (ashok@uthm.edu.my)

K S Tamilselvan

Department of Electronics and Communication Engineering, KPR Institute of Engineering and Technology, Coimbatore, India. (kstamilselvan@gmail.com)

Abstract: Seizure detections/predictions, classification for motor images, mental assignment classification, emotion, and sleep status classification, medication diagnosis, digital electroencephalographic (EEG) signal treatment were all highly popular in several applications. Effective algorithms must be employed in various methods for selecting a channel with a high number of EEG channels. The complex calculation of any EEG signal processing function is reduced by selecting the channels to be installed, thereby extracting critical features; (ii) reducing overfitting due to unneeded performance-enhancing channels; And the procedure for the channel selection is based on: signals such as time-domain analysis, spectrum power and wavelet modification are applied to extract characteristics and therefore to select the channel most channel selection methods.The present method has a drawback of large time delay.CEEG(conventional electroencephalogram) has several logistical limitations, including a short duration (typically under 60 minutes) and restricted access to adequately trained electrophysiologists. The selection of the EEG reduces the difficult calculation of any signal processing function channels to be installed.

Keywords: EEG signals; Channel range; Seizure identification; Sleep state categorization; Motor imagery categorization; Emotion categorization; Mental task categorization.