Improved performance on seizure detection in an automated electroencephalogram signal under evolution by extracting entropy feature

• Published in: Multimedia tools and applications Vol. 81; no. 10; pp. 13355 - 13370
• Main Authors: Munirathinam, Revathi, Ponnan, Suresh, Chakraborty, Chinmay, Umathurai, Saravanakumar
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.04.2022; Springer Nature B.V
• Subjects: 1176: Artificial Intelligence and Deep Learning for Biomedical Applications; Artificial neural networks; Automation; Computer Communication Networks; Computer Science; Convulsions & seizures; Data Structures and Information Theory; Electroencephalography; Epilepsy; Feature extraction; Human error; Multimedia Information Systems; Search algorithms; Seizures; Special Purpose and Application-Based Systems; Artificial neural network; Performance evolution; Seizure detection; EEG signal; Algorithm
• ISSN: 1380-7501; 1573-7721
• DOI: 10.1007/s11042-021-11069-7

In recent years, enormous people in all age category were affected with epilepsy throughout the world. To detect and evaluate the epilepsy seizure, the measurable component Electroencephalography (EEG) plays a major role in it. In the olden days, manual detection by the following medical process is carried, the accuracy of the detection is good, but some human error may occur which leads to the critical situation. The EEG plays a major role in the medical field to analyze the instant health condition of the subject under monitoring. The manual detection of EEG consumes much time and have to face the critical consequence, to avoid this situation world need alternate detecting methods. For more than a decade, to help medical specialists, many techniques, methodologies have been followed to detect with technology advancement. The detection of EEG signal was automated with various methods which reduce the detection time and gives the earlier response which is useful for medical diagnosis. During automated signaling the device also generates a noisy signal that causes difficulty in detection and prediction. In this paper, a new technique is proposed with an adaptive artificial neural network (AANN) to detect a normal and epileptic signal. The optimized proposed oppositional crow search algorithm (OCSA) gives better performance in the detection of epileptic seizures with an accuracy of 96.45% over others and the same has been reported.