Why Signal Quality Matters

An artifact in EEG is any recorded signal that does not originate from brain electrical activity. Because EEG amplifiers are designed to detect microvolt-level signals, they are highly sensitive to all electrical potentials near the electrodes — including those generated by muscle contractions, eye movements, cardiac rhythm, electrode issues, and environmental sources. If these non-cerebral signals are not identified and managed, they contaminate the data and can lead to inaccurate analysis.

Muscle artifacts (EMG contamination) are among the most common. Clenching the jaw, tensing the forehead, swallowing, or even subtle neck tension produces electrical potentials that are orders of magnitude larger than cortical signals. These artifacts typically manifest as high-frequency broadband noise, most visibly in the beta and gamma ranges, and can severely inflate power values in frontal and temporal channels. Eye movement artifacts (EOG) affect frontal electrodes — blinks produce sharp vertical deflections, while lateral eye movements create asymmetric slow-wave distortions across the front of the head.

Electrode-related artifacts arise from poor scalp contact, dried gel, loose sensors, or high impedance at one or more sites. These issues cause signal dropout, excessive noise, or drift that can mimic actual brain activity. A single bad electrode can distort an entire region of a topographic map. Similarly, 60 Hz line noise from power outlets and nearby electronics can inject a steady oscillation into the recording that masquerades as high-beta or gamma activity if not filtered appropriately.

Artifact management involves both prevention and post-processing. During acquisition, proper skin preparation, correct gel application, impedance checks, and instructing the client to relax their face and body minimize the introduction of artifacts. After recording, trained technicians visually inspect the data epoch by epoch, flagging or removing contaminated segments before quantitative analysis. Many processing pipelines also use automated artifact rejection algorithms, but manual review remains essential because automated methods can miss subtle contamination or incorrectly reject clean data.

When artifact contamination is severe or widespread — for example, if a client was unable to remain still, or multiple electrodes had poor contact — the affected segments or the entire recording may need to be discarded and re-recorded. This is not a failure; it is a quality assurance measure. An analysis built on artifact-laden data is worse than no analysis at all, because it produces results that look legitimate but do not reflect actual brain activity. At EEG Paradox Solutions, flagged data is always disclosed, and re-recording is recommended whenever data quality falls below acceptable thresholds.