Iterative Dichotomiser Maximum Posteriori Active Selection Algorithm for Analysis of Coma Patient’s Brain Waves Through WSN

• Published in: SN computer science Vol. 3; no. 3; p. 211
• Main Authors: Nazar, M. Mohamed Zamam, Surputheen, M. Mohamed
• Format: Journal Article
• Language: English
• Published: Singapore Springer Nature Singapore 01.05.2022; Springer Nature B.V
• Subjects: Algorithms; Automation; Biomarkers; Brain; Brain research; Coma; Communication; Computer Imaging; Computer Science; Computer Systems Organization and Communication Networks; Convulsions & seizures; Data Structures and Information Theory; Differentiation; Electroencephalography; Frequencies; Frequency ranges; Gabor filters; Information Systems and Communication Service; Interfaces; Internet of Things; Monitoring systems; Nervous system; Original Research; Patients; Pattern Recognition and Graphics; Physical properties; Remote sensors; Sensors; Software Engineering/Programming and Operating Systems; Statistical analysis; Vision; Voltage pulses; Wireless networks; Wireless sensor networks; Coma patient details; Gabor filter; Discrete Weight-Based Feature Selection (DW-FS); EEG signal data; ID-maximum posteriori Active Selection algorithm (ID-MPASA)
• ISSN: 2662-995X; 2661-8907
• DOI: 10.1007/s42979-022-01101-4

Coma is an unconscious state wherein the patient is unable to respond. An Electroencephalogram (EEG) is a test of the brain’s ability to record electrical activity. The nervous system of the brain consists of millions of neurons. Differentiation of the quantification and statistical analysis is difficult to signify the contributing features of the EEG signals. The proposed system-making process comprises three main stages: the first stage, a preprocessing-based Gabor Filter which removes ocular artifacts that affect the brain wave signal in the lower frequency range value analysis. In the next-stage, Discrete Weight-Based Feature Selection (DW-FS) method is implemented which efficiently selects the spectral frequency bands. It derives the feature based on density in each of these (delta band, theta and, alpha band, beta band, gamma-band) bands. The classification is the greatest effort focused on data representation stage, the potential of extracted feature set, designing, implementing and deciding an appropriate of the ID (Iterative Dichotomiser)-maximum posteriori Active Selection algorithm to enhance the differentiation and identification of coma states. Therefore, the IoT sensor can be used to monitor the patient’s coma stage for which effective signal data is considered. It records the activity of the brain by placing electrodes on the scalp to measure abnormal changes in voltage pulses generated by the brain, and by IoT transmitting and data collected for remote analysis in the wireless sensor network. These steps are using a Wireless Sensor Network (WSN), which is continuously monitored for a patient’s physical parameters in coma in real-time application.