From Neurons to Networks: A Holistic Review of Electroencephalography (EEG) from Neurophysiological Foundations to Advanced Decoding

Electroencephalography (EEG) has transitioned from a subjective observational method into a data-intensive analytical field that utilises sophisticated algorithms and mathematical modeling. This progression encompasses developments in signal preprocessing, artifact removal, and feature extraction techniques including time-domain, frequency-domain, time-frequency, and nonlinear complexity measures. To provide a holistic foundation for researchers, this review begins with the neurophysiological basis, recording technique and clinical applications of EEG, while maintaining its primary focus on the diverse methods used for signal analysis. It offers an overview of traditional mathematical techniques used in EEG analysis, alongside contemporary, state-of-the-art methodologies. Machine Learning (ML) and Deep Learning (DL) architectures, such as Support Vector Machines (SVMs), Convolutional Neural Networks (CNNs), and transformer models, have been employed to automate feature learning and classification across diverse applications. We conclude that the next generation of EEG analysis will likely converge into Neuro-Symbolic architectures, synergising the generative power of foundation models with the rigorous interpretability of signal theory.