EEGderived

EEGderived refers to metrics, features, and analyses derived from electroencephalography (EEG) data. In research and clinical contexts, EEG-derived measurements are used to quantify brain activity, detect abnormal patterns, or serve as inputs to predictive models. The term emphasizes the transformation of raw EEG signals into interpretable quantities.

Common EEG-derived features include spectral power in standard bands (delta, theta, alpha, beta, gamma), event-related potentials and their time–frequency representations, and metrics of functional connectivity such as coherence, phase-locking value, and phase-synchronization. Nonlinear and complexity-based measures, such as entropy and fractal dimensions, are also used to capture signal regularity and dynamic variability.

Processing typically involves data collection with scalp electrodes, preprocessing to remove artifacts (eye movements, muscle activity) using filtering and methods like independent component analysis, followed by feature extraction. Features may feed statistical tests or machine learning models for tasks like classification, diagnosis, or monitoring.

Applications span clinical and research domains, including epilepsy detection and monitoring, sleep staging, cognitive workload assessment, mood and neurodegenerative disorder research, and brain-computer interfaces. EEG-derived biomarkers are often sought for objective measures of brain state or pathology.

Challenges include inter-individual variability, sensitivity to preprocessing choices, and the need for standardized reporting. Initiatives such as EEG-BIDS promote consistent data organization and sharing to improve reproducibility and cross-study comparisons.