The EEG Eye Blink Artifact: What It Is and How to Fix It

An electroencephalogram, or EEG, is a tool that allows us to listen to the brain’s electrical conversations. This recording of brain activity can be disrupted by “artifacts,” which are signals from non-brain-related sources, much like static interfering with a radio broadcast. One of the most common and disruptive artifacts originates from a simple, everyday action: the blink of an eye.

The Biological Origin of Eye Blink Artifacts

An eye blink creates a significant electrical signal due to the eyeball’s natural electrical properties. The eye functions like a small biological battery, maintaining a steady electrical difference between its front and back. This is known as the cornea-retinal potential, where the cornea at the front of the eye is positively charged and the retina at the back is negatively charged.

During a blink, the eyelid moves down over the cornea. As the eyelid, a conductive surface, slides over the positively charged cornea, it effectively shunts this electrical charge across the skin of the forehead. This sudden change in the electrical field is powerful enough to be detected by the sensitive EEG electrodes on the scalp, particularly those on the forehead like Fp1 and Fp2.

The movement of the eyeball that accompanies a blink, known as Bell’s Phenomenon, also contributes. As the eyelid closes, the eyeball naturally rolls slightly upward. This rotation moves the positive pole (the cornea) closer to the frontal electrodes, further amplifying the electrical signal they pick up.

Recognizing the Artifact on an EEG

On an EEG display, the eye blink artifact has a distinct and recognizable appearance. It typically manifests as a high-amplitude, sharp wave that is positive in polarity. This signal appears synchronously, meaning at the same time, across the electrodes placed on the left and right frontal regions of the head. The amplitude of a blink artifact is often significantly larger, by several orders of magnitude, than the underlying brain waves being measured.

This signature allows it to be distinguished from other types of artifacts. For instance, a sideways eye movement, or saccade, produces a different pattern. A saccade typically appears as a slower, more square-shaped wave. It also shows an opposite polarity in electrodes on the sides of the head, such as F7 and F8; a glance to the left would create a positive wave at F7 and a negative one at F8.

Why Eye Blinks Complicate EEG Analysis

Eye blink artifacts in an EEG recording pose two challenges for analysis. The first is obscuration, as the artifact’s large voltage can completely mask the much smaller electrical signals produced by the brain that occur at the same moment. This means that potentially important neural data, such as subtle signs of seizure activity, can be hidden from view.

The second problem is the risk of misinterpretation. To an untrained observer, the sharp and spiky appearance of a blink artifact can look very similar to pathological brain signals, specifically epileptiform discharges which are markers for epilepsy. This resemblance creates a potential for diagnostic error if the artifact is not correctly identified and is instead mistaken for abnormal brain activity.

Methods for Artifact Removal

Several strategies exist to handle eye blink artifacts. The most straightforward approach is artifact rejection, which involves identifying and deleting the specific segments of the EEG data that are contaminated by blinks. While simple, the primary drawback of this method is the loss of all data recorded during that time, which may have contained valuable brain activity.

A more sophisticated technique involves using regression-based methods. This approach often requires placing dedicated electrooculogram (EOG) electrodes above and below an eye to get a clean recording of the blink itself. Mathematical algorithms use this EOG recording as a template to estimate and subtract the artifact’s influence from the corresponding EEG channels, cleaning the data without discarding it.

The most powerful and widely used method today is Independent Component Analysis (ICA). ICA is a computational signal processing technique that works like a filter for complex audio. Using a “cocktail party” analogy, ICA can listen to a mixture of signals—the brain’s activity, eye blinks, muscle noise—and separate them into their individual sources. Once the eye blink is identified as a distinct “component,” it can be mathematically removed, leaving the underlying brain signals intact for analysis.