Stages of Sleep: A classification analysed from Electroencephalogram Signals

Investigating the various stages of sleep, an essential aspect of human health and well-being, is crucial for understanding and improving overall health. A deep and detailed analysis of sleep stages is key to identifying sleep disorders and assessing sleep quality. Traditionally, sleep stage classification has relied on polysomnography, an advanced method that tracks various physiological indicators. However, the complexity and high cost of these systems limit their accessibility for widespread use. This paper introduces an accessible alternative for sleep stage classification by applying machine learning techniques to EEG data. It highlights a novel model that combines Convolutional Neural Networks (CNNs) and Long Short-Term Memory (LSTM) networks to effectively interpret the intricate temporal sequences in EEG signals, which are challenging to analyze with traditional methods. The outcomes of this research are promising, demonstrating that the model can identify different sleep stages with an accuracy exceeding 85%, outperforming conventional feature-based classifiers.