Analysing the Thought Process of EEG Pattern using Compressed Sensing Architecture

Sleep pattern recognition plays a crucial role in detecting pathological and psychological diseases. Various disorders can be identified through analysis of EEG patterns recorded during sleep. Sleep EEG consists of four primary waveforms: alpha, beta, theta, and delta waves, each associated with different sleep stages. The cyclic alternating pattern (CAP) is characterized by cerebral activity and autonomic motor functions, providing insights into motor events and neurovegetative functions that aid in understanding the pathophysiology of sleep disorders. This research focuses on identifying sleep patterns using a Compressed Sensing Architecture (CSA). The aim is to assist pathologists in accurately and efficiently diagnosing sleep disorders through automated analysis. Existing methodologies for extracting sleep patterns from EEG rely on various algorithms. In this study, error signals are extracted using CSA, and metrics such as the Percentage Root-mean-square Difference (PRD) and Signal-to-Noise Ratio (SNR) are computed after reconstructing the original signal. The proposed approach demonstrates enhanced accuracy, making it a promising solution for automated, error-free diagnosis of sleep disorders. The research findings have significant potential for practical implementation, improving diagnostic precision and clinical outcomes.